Here is the daily technology #threadcast for 10/22/24. We aim to educate people about this crucial area along with providing information of what is taking place.
Drop all question, comments, and articles relating to #technology and the future. The goal is make it a technology center.
Apple could roll out iOS 18.1 next week, starring ‘Apple Intelligence’
Apple Intelligence is coming soon. The Cupertino-headquartered tech titan is set to roll out its highly anticipated iOS 18.1 update sometime next week, bringing a host of AI-powered features to Apple’s family of devices.
As of now, the tech behemoth is yet to reveal an exact release date for iOS 18.1. Several reports however indicate that the update is expected to be available on October 28, marking the introduction of the first set of Apple Intelligence features. With this, users will have access to advanced AI capabilities, such as Writing Tools, notification summaries, updates to the Control Center, and a Clean Up feature in the Photos app. Some features will be introduced in the initial rollout, while the others will be introduced over the course of the coming year.
#technology #apple #ai
Speaking of features, iOS 18.1 is set to bring advanced Writing Tools to users, providing real-time aid in text compositions with suggestions for spelling, grammar, and even word choice. With this, users will be able to adjust the tone of their writing (as necessary) and will be able to choose between three distinct tones: Friendly, Professional, or Concise. Alongside this comes an option to summarize the notifications on devices, ensuring that the most critical alerts rise to the top and the less important ones are grouped together. This will be useful when you receive a plethora of notifications, and do not need to scramble across those to receive important updates.
Other upcoming features include an update to Siri – the assistant will get a completely new redesign in the form of a new user interface with a glowing light indicator, as well as the ability to maintain context throughout multiple interactions, amongst others. Users will also have the option to record their calls (this will be able available to iPhone XS models and newer), while the transcripts from the calls will be made available in English, Spanish, Mandarin, Cantonese, and Korean.
The launch of iOS 18.1 has been highly-anticipated, ever since Apple introduced iOS 18 last month alongside its latest addition to the iPhone lineup – the iPhone 16 series. Now, the company is set to roll out a new firmware upgrade for AirPods Pro 2 users as well – one that is slated to bring new hearing health features for users. For now, the specifics are yet to be finalized, but reports state that these will include hearing protection, hearing testing, and hearing aid functionalities (the device will be able to automatically adjust its audio output to act as hearing aids).
Reports also state that this update will let users conduct a hearing test directly from their iPhone itself – tones from multiple frequencies are utilized in this test. Once it is complete, users need to go to the Health app to see the results and adjust the audio settings as necessary.
Solar power from space? Actually, it might happen in a couple of years.
The idea of harvesting solar power continuously from space, where there are no nights, clouds, or atmosphere to interfere with the collection of photons, and beaming it back down to Earth via microwaves sounds simple, but implementing the technology is difficult. Aetherflux, a small company with only about 10 employees, plans to launch a satellite to test the idea and determine both the safety and efficiency of collecting solar energy in space. If the demo mission works, Aetherflux plans to develop a constellation of satellites to deliver power to remote locations.
#technology #solarpower #energy #space
Ray-Ban Meta Glasses Are The Top Selling Product In Many Ray-Ban Stores
Ray-Ban Meta glasses are the top-selling product in 60% of Ray-Ban stores across Europe, the Middle East, and Africa. The company's sales figures suggest that consumers are willing to pay more for smart glasses if they're stylish and widely available to try on. Meta and EssilorLuxottica have extended their partnership into the 2030s. The companies plan to continue developing multi-generational smart eyewear products.
#meta #metaglasses #rayban #EssilorLuxottica
ByteDance intern fired for planting malicious code in AI models
An intern at ByteDance maliciously interfered with the company's model training tasks, sabotaging its AI models. They were fired for committing serious disciplinary violations. None of the intern's sabotage impacted ByteDance's commercial projects or online business and none of the company's large models were affected. Rumors suggest the intern's actions cost the company tens of millions of dollars in losses, but ByteDance says those claims are exaggerated.
#technology #infosec #bytedance #ai
Elon Musk rolls out a controversial new $1Mn daily voter giveaway, and it is already under legal lens
Voters in Pennsylvania are now getting a unique incentive to vote in the upcoming elections – a lot of cash. As strange as this sounds, this is exactly what’s happening, and the money is coming from the pockets of Elon Musk. At a rally in Harrisburg, Pennsylvania, the billionaire announced that he will be giving $1 million every day to a randomly selected voter who signs a petition from his political action committee (PAC), America PAC
#technology #newsonleo
“I have a surprise for you,” Musk said at the rally, adding that the prize money will be made available “every day from now until the election.” The presidential election is on November 5. “I think think is kind of fun, and you know, it seems like a good use of money basically,” Musk added. The first cheque has already been given on Saturday, and another check was handed out on Sunday.
This giveaway is framed as a campaign to promote free speech and the right to bear arms, and the winner will be chosen at random. It should be noted that this giveaway is only open to registered voters (those who have signed a pro-US Constitution petition by the America PAC, and interested users need to go to the America PAC website to sign it). For now, it is restricted to Pennsylvania, Georgia, Nevada, Arizona, Michigan, Wisconsin and North Carolina. All of these are key states for the upcoming elections – Musk even called Pennsylvania the “linchpin” of the election.
And if this is not enough, America PAC is also offering several other smaller monetary rewards. These include bonuses of $100 bonuses for those who are signing the petition, as well as further payments if they refer the petition to others. for referrals. Critics argue that these incentives create a lottery-like system where voter registration becomes a gateway to potential financial gain.
Campaign finance lawyer Brendan Fischer warned that this initiative “approaches a legal boundary.” It is likely that this scheme could catch the eye of the federal authorities, who can can either prosecute Musk, file a civil lawsuit, or issue a warning to put an end to this giveaway. This initiative also undermines the elections themselves, especially when a prominent billionaire like Musk can simply leverage his worth to sway can simply leverage his worth to sway individuals to vote for certain parties or individuals. In this case, the individual is Donald Trump, the ex-US President who has now been endorsed by Musk.
While this giveaway is extremely lucrative for the winners, the bigger picture is less amusing. Pennsylvania Governor Josh Shapiro – during an appearance on NBC’s Meet the Press – described this development as “deeply concerning,” while critics are arguing that this development could violate US election law itself. The law prohibits giving out financial incentives for voting or voter registration, and that is exactly what Musk is doing by creating a link between the reward and the registration of new voters.
“I think there are real questions with how he is spending money in this race, how the dark money is flowing, not just into Pennsylvania, but apparently now into the pockets of Pennsylvanians. That is deeply concerning,” Shapiro said.
According to 52 U.S.C. 10307(c), it is illegal to offer or accept payment for voting or registering to vote. Rick Hasen, a professor at UCLA and an expert in election law, stated that the structure of Musk’s lottery “clearly” breaches this law. For reference, the law states that indivuduals who “pay or offer to pay or accept payment either for registration to vote or for voting” can be slammed with a hefty fine of $10,000 or even five years in prison.
AI
AI-generated influencers outperform humans in engagement, research reveals
According to research by HypeAuditor, personalities created by artificial intelligence are recording impressive engagement rates, reaching 6.5% – a number that far exceeds the average of human influencers, whose rates generally do not reach this level.
#newsonleo #ai #technology
Neysa, which provides enterprises with accelerated, scalable AI infra, nabs $30Mn in Series A from Nexus VP, others
Neysa, a generative AI upstart that helps enterprises rapidly deploy AI solutions at scale, has announced closing of a $30Mn Series A round. The round, which comes just months after its $20Mn seed round earlier this year, was led by existing investors NTTVC, Z47 (formerly Matrix Partners India) and Nexus Venture Partners. Neysa’s rapid $50Mn early fundraise in a year, follows the fundraise spree that AI startups have been witnessing globally, and is a sign that VC activity in AI space is picking up speed in India as well.
Since its seed fund earlier this year, Neysa has already made its Neysa Velocis platform generally available. Velocis enables on-demand access to high-performance computing infrastructure and Neysa claims large orders from companies across media and entertainment, service providers, software vendors, and government.
“This investment brings us closer to our vision of democratizing access to Gen AI and AI solutions for organizations across the world,” said Sharad Sanghi, Co-Founder and CEO of Neysa. “We’re thrilled with the progress we’ve made since our seed round. The general availability of Neysa Velocis and the onboarding of customers across key sectors validate our approach. By offering purpose-built, cost- effective AI-native solutions, we empower organizations to innovate confidently in a world that is being reshaped by AI.” The company aims to “democratizes AI adoption” with purpose-built platforms and services for AI-native applications and workloads.
Neysa currently offers three AI solutions to enterprises — Velocis (which is generally available), Overwatch and Aegis. Velocis helps in accelerated deployment of AI products, providing MLOps & GPU Cloud to deploy, train and infer Gen AI and AI models out of the box. Overwatch is an AI powered network monitoring tool, that provides intelligent real-time monitoring and alerts with remediation via AI interpretation and insights across network. Aegis is a cyber security platform that provides protection against AI-specific attacks globally.
A NASSCOM report had pegged India’s AI demand to reach nearly $22Bn by 2027, growing at a CAGR of 25-35%. This, coupled with a high engineering talent pool, has led to the birth of several fast-growing AI upstarts in the country, with a slew of them raising multi-million VC investments amid global AI frenzy.
Qualcomm launches new Snapdragon 8 Elite chip for premium smartphones, focus on AI
Chipmaker Qualcomm, at its annual Snapdragon Summit in Hawaii, has now revealed the Snapdragon 8 Elite, its latest flagship mobile chip. The company describes it as “the most powerful and world’s fastest mobile system-on-a-chip ever,” and is set to come to high-end smartphones from leading manufacturers such as Samsung, OnePlus, Motorola, and Xiaomi in the coming weeks.
#technology #ai #smartphones
AI remains the focal point of the Snapdragon 8 Elite, and Qualcomm aims to use this chip to take generative AI on mobile phones to new heights. The company notes that the chip will make “on-device multi-modal generative AI applications a reality on smartphones,” and comes with a real-time camera assistant that can recognize objects as users capture photos or record videos. Alongside this comes support for up to 320MP camera sensors and AI-driven tools that are integrated into the image signal processor (ISP) to allow real-time for camera adjustments. The Snapdragon 8 Elite also builds upon Qualcomm’s previous breakthroughs in AI and is set to handle text, audio, and visual data simultaneously to provide answers to prompts on-device.
With the Snapdragon 8 Elite, Qualcomm is also shifting from Kryo CPUs to its custom-designed Oryon CPU. This is fabricated using a 3-nanometer process and promises to bring a rise of 45% in overall performance and a boost of 44% in power efficiency. In the chip, there are two two prime cores clocked at 4.32 GHz and six performance cores reaching up to 3.53 GHz, and the company claims that they provide an increase of 62% increase in web performance. In addition to this, the Adreno GPU has undergone a revamp to gain a “sliced” architecture, and will support Unreal Engine’s Chaos Physics system. The chip will be a boon for gamers as well, and they will experience smoother gameplay with higher framerates, as well as a longer gaming time by as much as 2.5 hours.
We are so excited to bring the power of Qualcomm Oryon to our Snapdragon mobile platforms for the first time. Earlier this year we debuted it in PCs, delivering remarkable experiences and unparallel battery life to PC users, energizing the industry and getting the attention of consumers. Today, our second generation of the Qualcomm Oryon CPU debuts in our flagship Mobile Platform – it’s a major leap forward and we expect consumers to be thrilled with the new experiences enabled by our CPU technology” Chris Patrick, Senior Vice President and General Manager of mobile handsets at Qualcomm, commented on the matter.
The Snapdragon 8 Elite also comes with an upgraded Hexagon neural processing unit (NPU), which is said to offer up to 45% faster AI performance than its predecessor. The NPU is capable of processing more than 70 tokens per second, and comes with support for on-device multimodal AI assistants. “With leading CPU, GPU and NPU capabilities, the Snapdragon 8 Elite delivers dramatic performance enhancements and power efficiency 3. In addition, it revolutionizes mobile experiences by offering personalized, multi-modal generative AI directly on the device enabling the understanding of speech, context, and images to enhance everything from productivity to creativity tasks while prioritizing user privacy,” Patrick added.
https://inleo.io/threads/view/ciskodisco.opus/re-leothreads-29zfpdjr6
There's Something Weird About ChatGPT o1 Use Cases...
Summary below ⏬
Conclusion
While o1 represents an impressive advancement in AI technology, particularly in its approach to inference-time computation, the analysis suggests that many organizations might find GPT-4o sufficient for their current needs. The decision to adopt cutting-edge models should be based on specific use cases that genuinely require advanced reasoning capabilities beyond what existing frontier models can provide.
Hi, @mightpossibly,
This post has been voted on by @darkcloaks because you are an active member of the Darkcloaks gaming community.
Get started with Darkcloaks today, and follow us on Inleo for the latest updates.
GPT-4o vs o1: Are Cutting-Edge Models Necessary for Most Use Cases?
A recent analysis of OpenAI's o1 model demonstrations reveals an intriguing insight about the practical necessity of cutting-edge AI models. While OpenAI showcased o1's capabilities across various use cases, subsequent testing suggests that GPT-4o can handle many of these tasks just as effectively, raising questions about when organizations truly need the latest AI advancements.
The 90-8-2 Rule of AI Use Cases
According to the analysis, AI use cases generally fall into three tiers:
Key Differences Between GPT-4o and o1
The primary distinction between GPT-4o and o1 lies in their processing approach:
It's worth noting that o1 currently has some limitations compared to GPT-4o:
Testing o1's Core Use Cases
OpenAI identified three primary areas where o1 supposedly adds the most value:
1. Strategy
A market expansion scenario was tested, where both models were asked to analyze potential office locations. The test revealed that GPT-4o produced comparable analysis of markets like Paris and Berlin, including considerations for:
2. Coding
Using o1 Mini (optimized for coding tasks), the test examined creating a web application with Node.js backend and React frontend. While o1 Mini demonstrated impressive speed and detailed step-by-step guidance, GPT-4o proved equally capable of:
3. Research
A practical research scenario involving dog food optimization was tested. Both models successfully:
Mathematical Capabilities
While traditional language models have struggled with mathematical tasks, both o1 Mini and GPT-4o demonstrated accurate handling of complex calculations, as shown in a covered call option analysis example. Both models arrived at identical conclusions regarding maximum profit calculations and opportunity costs.
Key Takeaways
The analysis suggests several important conclusions:
Capability Overlap: For many showcased use cases, GPT-4o demonstrated comparable performance to o1, raising questions about the necessity of upgrading for these specific applications.
Speed Considerations: o1 Mini showed impressive speed, particularly in coding tasks, potentially offering an advantage in scenarios where rapid response times are crucial.
Cost-Benefit Analysis: Organizations should carefully evaluate whether their use cases truly require o1's enhanced capabilities, given the potential cost implications.
Future Potential: While current use cases might not fully demonstrate o1's advantages, the model's innovative approach to inference-time computation scaling could prove valuable as the technology evolves.
xAI, Elon Musk's AI startup, launches an API
Elon Musk's AI startup, xAI, has launched an API. But it's a bit barren at the moment.
#xai #grok #ai #technology
xAI's Grok API debuts with Limited Functionality and Technical Issues
In a significant development, xAI, a company founded by Elon Musk, has launched its flagship generative AI model, Grok, as an API (Application Programming Interface). The API allows developers to integrate Grok's capabilities into their own applications, but its initial release is somewhat bare-bones and plagued by technical issues.
The xAI API currently offers a single model, "grok-beta," which is priced at $5 per million input tokens (approximately 750,000 words) or $15 per million output tokens. Tokens are subdivided bits of raw data, such as syllables or characters. However, it is unclear which Grok model "grok-beta" is, as the documentation refers to Grok 2 and Grok mini, a lightweight and more affordable version of Grok.
Some users on X, the platform where Grok is integrated, have reported encountering problems paying for usage credits. The API supports function calling, which enables Grok models to connect with external tools such as databases and search engines. Additionally, the documentation hints at the possibility of vision models capable of analyzing both text and images, although these features are not yet live.
Grok, the AI model, has gained notoriety for its "rebellious streak," which allows it to answer questions that other AI systems might reject. For example, if asked to be vulgar, Grok will happily oblige, generating profanities and colorful language. This feature has raised concerns about the potential misuse of the model.
Grok has become an integral part of the X experience, with capabilities such as generating images using the open image generator Flux, summarizing news and trending events, and enhancing search capabilities, account bios, post analytics, and reply functions.
Despite the initial limitations, xAI is racing to catch up with competitors like OpenAi and Anthropic in the generative AI space. In May, the company raised $6 billion in a funding round led by investors including Andreessen Horowitz, Sequoia capital, and Fidelity. xAI's pitch for the funding round included a vision where its models would be trained on data from Musk's various companies, including Tesla, SpaceX, and The Boring Company, and would improve technology across those companies.
However, Tesla shareholders have expressed concerns about the diversion of resources and talent from Tesla to xAI. Several shareholders have sued Musk over the decision to start xAI, arguing that it is a competing venture that undermines Tesla's interests.
In related news, xAI is training the next generation of Grok models at its Memphis data center, which has been accused of worsening smog with unauthorized turbines. The company hopes to upgrade the data center next year, but will need approval from the Tennessee Valley Authority to do so.
The launch of the xAI API marks an important milestone for the company, but its limited functionality and technical issues may hinder its adoption in the short term. As xAI continues to develop its capabilities and address these issues, it will be interesting to see how it competes with established players in the generative AI space.
Some of the key features of the xAI API include:
Overall, the launch of the xAI API is a significant development in the generative AI space, but its limitations and technical issues may hinder its adoption in the short term. As xAI continues to develop its capabilities and address these issues, it will be interesting to see how it competes with established players in the market.
Article
What is an API?
An application Programming Interface (API) is a crucial tool that enables different applications, services, or systems to communicate with each other by providing a set of defined rules. This allows data to be shared, processed, and used across various platforms, languages, and systems, fostering a seamless exchange of information.
When a company sets up an API, it essentially creates a gateway that allows other developers to access its data, services, or functionality, enabling them to build applications, services, or integrations that interact with the company's system without having to build everything from scratch. This process involves several key steps:
By setting up an API, a company can achieve several benefits, including:
Examples of APIs include:
In summary, an API is a set of rules that enables different systems to communicate with each other, allowing data and functionality to be shared and used across different platforms and services. By setting up an API, a company can open up new opportunities for innovation, collaboration, and revenue generation, while also improving its customers' experience and expanding its reach.
Physicists Generated Sound Waves That Travel in One Direction Only
Imagine three people huddled in a circle so when one speaks, only one other hears. Scientists have created a device that works like that, ensuring sound waves ripple in one direction only.
The device, developed by scientists at ETH Zurich and the Swiss Federal Institute of Technology Lausanne, is made up of a disk-shaped cavity with three equally-spaced ports that can each send or receive sound.
#science #soundwaves
In an inactive state, sound transmitted from port 1 is audible to ports 2 and 3 at equal volumes. Sound waves bounce back to port 1 as an echo as well.
When the system is running, however, only port 2 hears port 1's sounds.
The trick is to blow swirling air into the cavity at a specific speed and intensity, which allows the sound waves to synchronize in a repeating pattern. That not only guides the sound waves in a single direction, but gives more energy to those oscillations so they don't dissipate. It's kind of like a roundabout for sound.
Article
Yesterday we had over 100 articles and stories linked in the technology threadcast.
This is becoming similar to a subreddit.
Poshmark reverses decision to lower selling fees after sellers experience a drop in sales
Shortly after the social marketplace Poshmark announced it would lower selling fees and increase fees for buyers, the company decided to revert to its Shortly after Poshmark lowered selling fees and increased fees for buyers, it reverted back to the original fee structure following outrage among sellers.
#poshmark #sellers #fees #newsonleo
Shortly after the social marketplace Poshmark announced it would lower selling fees and increase fees for buyers, the company decided to revert to its original fee structure following outrage among sellers.
In an email to sellers sent on Monday (and viewed by TechCrunch), founder and CEO Manish Chandra wrote an apology about the fee change, saying it resulted in shoppers spending less on purchases, “leaving our sellers with less cash in their pockets—despite the seller fee reduction.”
Article
These countries are leading the way to 100% renewable electricity
By 2030, 10 countries are expected to reach or surpass the 66% wind and solar penetration level.
At least 10 countries are on track to get more than two-thirds of their electricity from wind and solar by 2030, according to the International Energy Agency.
#energy #renewable #technology
A handful of nations have already reached — or come close to — the 100% renewables mark, including Norway, New Zealand, Iceland, Costa Rica, Kenya, Bhutan, and Paraguay. However, they relied on traditional technologies — including hydro and first-generation geothermal power plants — and most countries don’t have access to these resources in abundance.
Some others, including France, get most of their electricity from low-carbon nuclear power.
The IEA’s 2024 renewable energy stocktake focuses partly on the countries that are leading the shift to variable renewable energy technologies — primarily wind and solar, which are now the cheapest sources of generating capacity in much of the world.
Denmark is the current frontrunner, and is on track to get within touching distance of 100% renewables by the end of the decade, the IEA projects.
Article
CrewAI uses third-party models to automate business tasks
CrewAI, a startup developing a hub where companies can build automations using third-party AI models, has raised $18 million from investors including Andrew Ng.
Back in 2022, João Moura was directing AI engineering efforts at Clearbit, a startup creating a unified hub for business intelligence tools. There, Moura was responsible for leading the development of AI integrations, as well as defining Clearbit’s AI product roadmap.
#crewai #technology #newsonleo #andrewng
CrewAI: Revolutionizing Workflow Automation with AI
In today's fast-paced business environment, workflow automation has become a crucial aspect of operations to streamline processes, increase efficiency, and drive growth. CrewAI, a cutting-edge startup, is leading the charge in this space by harnessing the power of artificial intelligence (AI) to automate repetitive, back-office tasks. In this article, we'll delve into the world of CrewAI, exploring its founding story, innovative platform, and future plans.
The Founding Story
João Moura, a seasoned entrepreneur with a background in AI engineering, founded CrewAI after leaving his role as the director of AI engineering at Clearbit, a startup that was acquired by HubSpot. Moura's previous venture, Urdog, developed a smart collar for pets, but this time, he had a more ambitious vision in mind. CrewAI aims to automate tasks such as summarizing reports, onboarding employees, and enriching marketing databases using AI models from vendors like OpenAi and Anthropic.
The Platform
CrewAI's platform offers a user-friendly interface for customers to build workflow automations, deploy and track them from a dashboard, and integrate with over 1,000 different applications. Unlike traditional robotic process automation (RPA), which relies on rigid "if-then" rules, CrewAI leverages existing AI models to drive automations. This approach enables businesses to automate tasks such as analyzing customer feedback, forecasting trends, and protecting data privacy.
A Game-Changer in Workflow Automation
CrewAI's approach is a departure from traditional RPA, which can be inflexible and prone to errors. Moura argues that AI is a more resilient technology, capable of adapting to changing business needs. With CrewAI, teams can build groups of AI "agents" to perform tasks, integrate with various applications, and protect data privacy.
Competition and Funding
CrewAI has attracted significant funding, raising $18 million across seed and Series A rounds from prominent investors, including Boldstart Ventures, Craft Ventures, Earl Grey capital, and Insight Partners. The company has also secured investments from notable entrepreneurs, such as Andrew Ng, co-founder of Coursera, and Dharmesh Shah, co-founder and CTO of HubSpot.
While CrewAI faces competition from other AI-powered workflow automation startups, such as Orby, Bardeen, Tektonic, 11x.ai, Twin Labs, and Emergence, the company has managed to attract a sizeable customer base of 150 businesses in its first year.
Enterprise Cloud and Future Plans
CrewAI has launched Enterprise Cloud, a managed subscription plan built on tOP of open-source components, providing additional access controls, analytics, and VIP support. The company plans to use its funding to grow its 16-person workforce and expand its core automation products.
Conclusion
CrewAI is revolutionizing the workflow automation landscape by leveraging AI to automate repetitive tasks. With its innovative approach, the company has attracted significant funding and a growing customer base. As the company continues to expand its offerings and workforce, it's poised to become a major player in the industry.
Article
Anthropic just made it harder for AI to go rogue with its updated safety policy
Anthropic, the artificial intelligence company behind the popular Claude chatbot, today announced a sweeping update to its Responsible Scaling Policy (RSP), aimed at mitigating the risks of highly capable AI systems.
#technology #anthropic #ai #safety #newsonleo
The policy, originally introduced in 2023, has evolved with new protocols to ensure that AI models, as they grow more powerful, are developed and deployed safely.
This revised policy sets out specific Capability Thresholds—benchmarks that indicate when an AI model’s abilities have reached a point where additional safeguards are necessary.
Article
Microsoft’s new AI agents set to shake up enterprise software, sparking new battle with Salesforce
Microsoft just announced a suite of autonomous AI agents for its Dynamics 365 platform, intensifying the competition with Salesforce in the enterprise AI market. The tech giant will release ten new autonomous agents designed to augment sales, service, finance, and supply chain teams.
#microsoft #ai #agents #technology #enterprise
Microsoft Unveils Autonomous AI Agents for Dynamics 365, Intensifying Competition with Salesforce
In a move that is set to revolutionize the enterprise AI market, Microsoft has announced a suite of autonomous AI agents designed to augment sales, service, finance, and supply chain teams within its Dynamics 365 platform. This strategic move comes just weeks after Salesforce unveiled its Agentforce platform, and is seen as a direct challenge to the CRM giant's offering.
The ten new AI agents, available in public preview starting next month, are designed to automate complex tasks and orchestrate business processes across organizations. Unlike traditional chatbots and Microsoft's earlier AI offerings, these agents are capable of reasoning over intent and context, making judgments based on a broader set of data. According to Bryan Goode, Corporate Vice President at Microsoft, these agents are not meant to replace human workers but to enhance their capabilities.
"We think of these agents as really the apps of the AI era," Goode said in an interview with VentureBeat. "Every line of business system that exists today is going to get reimagined as an agent that sits on tOP of a copilot."
Microsoft's move is seen as a direct response to Salesforce's Agentforce platform, which was unveiled earlier this year. Salesforce CEO Marc Benioff has been aggressively promoting Agentforce, while criticizing Microsoft's Copilot. Benioff recently referred to Microsoft Copilot as "more like Clippy 2.0," a nod to Microsoft's much-maligned Office assistant from the 1990s.
However, Microsoft's agents are powered by advanced language models and the company's vast troves of enterprise data, giving it a unique advantage over Salesforce's Agentforce, which relies on its Atlas reasoning engine.
The tech industry is witnessing a paradigm shift as AI agents move from experimental technology to core business tools. Microsoft and Salesforce are at the forefront of this shift, each leveraging their unique strengths to shape the future of enterprise software.
Microsoft's strategy hinges on its ubiquitous presence in office productivity and cloud computing, making it easier for its vast user base to adopt AI agents. Salesforce, on the other hand, is banking on its CRM expertise and the power of its recently developed Data cloud to create AI agents that understand and optimize customer relationships.
The success of these platforms could redefine the future of work and enterprise software. As AI agents become more sophisticated, they may blur the lines between human and machine tasks, potentially reshaping organizational structures and job roles.
However, challenges remain. Data privacy concerns, the need for transparent AI decision-making, and the potential for job displacement are issues both companies must navigate carefully. Their ability to address these concerns while delivering tangible business value will likely determine the pace and extent of AI agent adoption.
The battle for AI dominance is heating up, and the outcome will have far-reaching implications for the future of enterprise software and the workforce. As the AI revolution unfolds, one thing is clear: the enterprise software landscape is on the cusp of a major transformation. Whether it's Microsoft's vision of "agents plus copilot plus humans" or Salesforce's "human at the helm" approach, the future of work is being rewritten — one AI agent at a time.
Key details:
Article
IBM debuts open source Granite 3.0 LLMs for enterprise AI
Make no mistake about it, enterprise AI is big business, especially for IBM.
IBM already has a $2 billion book of business related to generative AI and it’s now looking to accelerate that growth. IBM is expanding its enterprise AI business today with the launch of the third generation of Granite large language models (LLMs). A core element of the new generation is the continued focus on real open source enterprise AI. Going a step further, IBM is ensuring that models can be fine-tuned for enterprise AI, with its InstructLab capabilities.
#ibm #llm #enterprise #technology #granite
The new models announced today include general purpose options with a 2 billion and 8 billion Granite 3.0. There are also Mixture-of-Experts (MoE) models that include Granite 3.0 3B A800M Instruct, Granite 3.0 1B A400M Instruct, Granite 3.0 3B A800M Base and Granite 3.0 1B A400M Base. Rounding out the update, IBM also has a new group with optimized guardrail and safety options that include Granite Guardian 3.0 8B and Granite Guardian 3.0 2B models. The new models will be available on IBM’s watsonX service, as well as on Amazon Bedrock, Amazon Sagemaker and Hugging Face.
“As we mentioned on our last earnings call, the book of business that we’ve built on generative AI is now $2 billion plus across technology and consulting,” Rob Thomas, senior vice-president and chief commercial officer at IBM, said during a briefing with press and analysts. “As I think about my 25 years in IBM, I’m not sure we’ve ever had a business that has scaled at this pace.”
Article
Gartner: 2025 will see the rise of AI agents (and other top trends)
The pace of AI continues to accelerate, with capabilities never before thought possible now becoming a reality. This is particularly true of AI agents, or virtual co-workers, which will work alongside us and, eventually, autonomously.
#ai #technology #gartner #aiagents
In fact, Gartner predicts that by 2028, at least 15% of day-to-day work decisions will be made autonomously through agentic AI (up from 0% in 2024). Further emphasizing the technology’s potential, the firm has named it a top strategic technology trend in 2025.
“It’s happening really, really fast,” Gene Alvarez, distinguished VP analyst with Gartner, told VentureBeat. “Nobody ever goes to bed at night with everything done. Organizations spend a lot of time monitoring things. The ability to create agents to not only do that monitoring but take action will help not just from a productivity perspective but a timing perspective.”
The Cool and Scary Future of AI: Trends and Predictions
Artificial intelligence (AI) is revolutionizing the way we live and work, and its impact is only expected to grow in the coming years. From mundane tasks to complex scenarios, AI agents are becoming increasingly sophisticated, raising both excitement and concerns about their potential effects on society. In this article, we'll delve into the latest trends and predictions in AI, from its applications in the workforce to its potential impact on our brains.
Agentic AI: The Next Level of Automation
Agentic AI has the ability to plan, sense, and take action, making it a game-changer in various industries. For instance, AI agents can help upscale the workforce by guiding new employees and teaching them new skills. This could lead to a more efficient and productive workforce, but it also raises concerns about job loss and the potential for AI to displace human workers. As AI becomes more prevalent, it's essential to address these concerns and develop strategies to mitigate the negative impacts on employment.
Building Trust in AI
As AI becomes more pervasive, building trust in its capabilities is crucial. AI governance platforms will enable organizations to manage their AI systems' legal, ethical, and operational performance, ensuring transparency and responsible use. These platforms will also check for bias and provide information on how models were built, as well as the reasoning behind their prompts. This transparency is essential for building trust in AI and ensuring that it is used in a way that benefits society as a whole.
Hybrid Computing Models: The Future of Computing
The future of computing will involve hybrid models that combine different compute, storage, and network mechanisms. Orchestration software will move compute from one to another depending on the task and the method most suited for the job. This will require new technologies that use significantly less energy, as there is increased pressure to reduce consumption and carbon footprints. The development of more energy-efficient computing models will be critical in reducing the environmental impact of AI.
Disinformation Security: The New Frontier
AI is allowing threat actors to spread disinformation faster and more easily than ever before. Disinformation security is critical, and emerging technologies will assess authenticity, track the spread of harmful information, and prevent impersonation. By 2028, Gartner predicts that half of aLL enterprises will begin adopting products, services, or features specifically designed for disinformation security. As AI becomes more sophisticated, it's essential to develop effective strategies for detecting and preventing disinformation.
Preparing for the Post-Quantum World
quantum computing is rapidly advancing, and it will soon be able to break public key cryptography, or asymmetrical encryption. red teams are already harvesting encrypted data and holding onto it until quantum is realized. Organizations must be prepping for post-quantum cryptography nOW to ensure that their data is resistant to decryption. The development of post-quantum cryptography will be critical in protecting sensitive information and preventing data breaches.
AI Enhancing Our Brains
Gartner anticipates a rise in the use of bidirectional brain-machine interfaces (BBMIs) that read and decode brain activity and enhance human cognitive abilities. These could be directly integrated into our brains or made possible via wearables such as glasses or headbands. By 2030, 30% of knowledge workers will be using technologies such as BBMIs to stay relevant in the AI-powered workplace. The potential benefits of AI-enhanced brains include improved cognitive abilities, enhanced creativity, and increased productivity. However, it's essential to address the ethical implications of AI-enhanced brains and ensure that they are used in a way that benefits society as a whole.
Conclusion
The future of AI is both exciting and unsettling. As AI agents become more sophisticated, they will have the potential to transform various industries and aspects of our lives. However, it is essential to address the concerns and challenges that come with AI, such as job loss, disinformation, and cybersecurity threats. By building trust in AI, preparing for the post-quantum world, and enhancing our brains, we can harness the power of AI to create a better future for all. It's essential to prioritize responsible AI development and ensure that AI is used in a way that benefits society as a whole.
Dotmatics aims to speed drug development, break data silos with Geneious Luma
Even as enterprises across sectors explore how to integrate generative AI, it’s clear that biomedical research and the sciences are among the areas that could benefit most — as highlighted in the recent Nobel Prizes in Chemistry and Physics awarded to AI researchers.
#drugs #technology #bigpharma #dotmatics
Now Dotmatics, a leader in research and development scientific software, wants to give drug researchers the power of AI to speed up their development of new, life-saving and improving drugs.
Today, the Boston-based company launches Geneious Luma, a powerful new bioinformatics solution for antibody discovery, built on its Luma Scientific Intelligence Platform.
Geneious Luma is designed to streamline the process of biologic therapeutic discovery by integrating advanced sequence analysis, data management, and AI-powered automation. This release marks the first of several multimodal solutions aimed at transforming scientific research and accelerating the discovery of new therapies.
Thomas Swalla, CEO of Dotmatics, explained the benefit of Geneious Luma in a recent video interview with VentureBeat: “The problem we’re solving is that discovering new molecules and drugs has become increasingly complicated. It takes over a decade to bring a drug to market and costs anywhere from two to six billion dollars.”
Article
Here is a rewritten version of the text with more details:
Dotmatics, a software company founded in 1991, has been a pioneer in providing laboratory information management system (LIMS) and data management solutions for the life sciences industry.
The company's journey began with a team of scientists and software developers who recognized the need for a more efficient and effective way to manage laboratory data. Their first product, Dotmatics LIMS, was launched in 1994 and quickly gained popularity among pharmaceutical and biotechnology companies due to its ability to streamline laboratory operations, track samples, and analyze data.
Over the years, Dotmatics has continued to innovate and expand its product portfolio by acquiring several companies and developing new solutions to meet the evolving needs of the life sciences industry. Today, the company offers a range of software solutions that help life sciences organizations manage their laboratory data, streamline their workflows, and improve their overall efficiency.
At the heart of Dotmatics' offerings is its flagship product, Dotmatics LIMS, a comprehensive laboratory information management system that enables organizations to manage their laboratory operations, track samples, and analyze data. This solution provides a centralized platform for managing laboratory data, including sample tracking, data storage, and retrieval. Additionally, Dotmatics LIMS automates many laboratory workflows, reducing manual errors and increasing efficiency.
Beyond LIMS, Dotmatics offers a range of other solutions designed to meet the specific needs of life sciences organizations. These include:
The key features and benefits of Dotmatics' solutions include:
Overall, Dotmatics is a leading provider of laboratory information management system (LIMS) and data management solutions for the life sciences industry, helping organizations to manage their laboratory data, streamline their workflows, and improve their overall efficiency.
The AI edge in cybersecurity: Predictive tools aim to slash response times
Modern cybersecurity professionals require advanced technologies to deter, detect and expel hackers, and the predictive benefits of AI can mean the difference between data protection and ruin.
#cyberscurity #ai #security #technology
The average cost of a data breach in the U.S. hit a high-water mark of $9.48 million in 2023. Losses have ticked up every year since 2013, even during the global health emergency of Covid-19 when many businesses shuttered. An analysis in IBM’s 2024 data breach report indicates that organizations that employed extensive AI security automation saved $2.22 million, while also lowering cybersecurity insurance.
Industry leaders would be well-served to think about cyberattacks outside the financial implications, as well. Should your organization pay a ransomware demand or right the ship after a crushing malware attack, the reputational damage can far outweigh the dollars. When hackers steal confidential, sensitive and personal identity information, those in your orbit are negatively impacted. Employees, customers and industry partners may file civil actions.
Article
Zomato to raise $1 billion ahead of rival Swiggy India IPO
Zomato on Tuesday said that its board has approved a plan to raise $1 billion via a share sale to institutional investors.
Zomato on Tuesday said its board has approved a plan to raise $1 billion via a sale of shares to institutional investors, as competition in India’s quick-commerce sector heats up.
#zomato #india #ipo
The timing of the share sale, the company’s first major fundraise since its 2021 IPO, is notable, because it comes just weeks before its rival Swiggy’s public debut. Bengaluru-based Swiggy, backed by Prosus Ventures, SoftBank, and Accel, aims to raise about $1.4 billion in its IPO next month.
The move by Zomato, which one investor characterized as “sucking the air out of the room” for competitors, has caught market observers off-guard. Jefferies analysts called the capital raise “unexpected,” given Zomato’s existing $1.2 billion cash reserves. They suggest the strategy may be aimed at reducing foreign institutional investor ownership in Zomato below 50%, effectively transforming it into a “domestic” company.
The shift to majority domestic ownership could enable Blinkit, Zomato’s quick-commerce subsidiary, to adopt an inventory-based model in India. Current regulations restrict foreign-owned companies to conducting only marketplace operations, and prohibit them from owning inventory sold within the country.
Article
What is Zomato?
Here is a rewritten version with more details:
Zomato, a pioneering online food delivery and restaurant discovery platform, was founded in 2008 by Deepinder Goyal and Pankaj Chaddah in New Delhi, India. The company's journey began with a humble start, driven by the founders' passion for food and technology.
Early days (2008-2010)
Deepinder Goyal, a computer science graduate from the Indian Institute of Technology (IIT) Delhi, and Pankaj Chaddah, an engineering graduate from IIT Kanpur, met while working at a software company. They shared a common interest in food and technology, which led them to start a food review website called Foodiebay.com.
The website allowed users to review and rate restaurants, providing a comprehensive directory of eateries in the city. Foodiebay.com quickly gained popularity, with users sharing their dining experiences and recommendations.
Rebranding and Expansion (2010-2012)
In 2010, the founders decided to rebrand Foodiebay.com as Zomato.com, a name inspired by the Italian word "zomato," meaning "to eat." This marked a significant milestone in the company's history, as it expanded its features to include restaurant listings, menus, and user reviews. Zomato also launched its first mobile app for Android and iOS devices, making it easier for users to access the platform on-the-go.
International Expansion (2012-2015)
Zomato's rapid growth and success led to its expansion into international markets. The company entered the Middle East, Southeast Asia, and Latin America, establishing a strong presence in these regions. This expansion not only increased Zomato's user base but also provided a platform for local restaurants to showcase their offerings to a global audience.
Series A and B Funding (2012-2015)
Zomato's growth and expansion required significant funding to support its operations and develop new features. In 2012, the company raised Series A funding of $1 million from InfoEdge, a leading Indian Internet company. This funding enabled Zomato to expand its team and develop its technology infrastructure. In 2013, Zomato raised Series B funding of $25 million from Sequoia capital and InfoEdge, which further accelerated its growth and expansion plans.
Series C and D Funding (2015-2017)
Zomato's success in international markets and its growing user base led to further funding rounds. In 2015, the company raised Series C funding of $60 million from Sequoia Capital, InfoEdge, and other investors. This funding enabled Zomato to expand its food delivery service and enter new markets. In 2017, Zomato raised Series D funding of $200 million from Ant Financial, a Chinese e-commerce giant, which further solidified its position in the market.
IPO and Expansion (2018-Present)
Zomato's growth and success culminated in its initial public offering (IPO) in July 2021, raising $1.1 billion. The company used this funding to expand its operations, develop new features, and enter new markets. Today, Zomato is one of the largest food delivery and restaurant discovery platforms in India, with operations in over 10,000 cities across 24 countries.
Key Milestones
Throughout its journey, Zomato has demonstrated its ability to innovate, expand, and adapt to changing market conditions. Today, the company is a leading player in the Indian food delivery and restaurant discovery space, with a strong presence in international markets.
TikTok owner fires intern who 'maliciously interfered' with AI project
ByteDance confirmed the intern’s dismissal, writing on a Chinese social news platform.
TikTok’s China-based parent company ByteDance fired an intern for allegedly sabotaging a training session about artificial intelligence.
#tiktok #technology #bytedance #AI #sociamedia
ByteDance confirmed the worker was canned in August after having “maliciously interfered” on a research project to train AI models.
The company denied reports that the intern cost it “tens of millions of dollars,” saying that those claims were “seriously exaggerated.”
“The intern involved maliciously interfered with the model training tasks of the commercial technology team’s research project, but it did not affect the formal commercial projects and online business, nor did it involve other businesses such as ByteDance’s large models,” the TikTok owner wrote in the post on a Chinese social news platform.
Article
Senators Want Biden Admin Review of Undersea Cable Vulnerabilities
More than 95% of international internet traffic travels on undersea cables and the locations of the cables are often openly published to prevent accidental damage.
A bipartisan group of eight U.S. senators on Monday urged President Joe Biden to order a review of the security of the global network of undersea communications cables that handle nearly all the world's internet traffic.
#technology #underwatercable #security
More than 95% of international internet traffic travels on undersea cables and the locations of the cables are often openly published to prevent accidental damage.
"It is imperative that the United States undertake a review of existing vulnerabilities to global undersea cable infrastructure, including the threat of sabotage by Russia as well as the growing role of the People’s Republic of China in cable laying and repair," said the letter, led by Republican Todd Young and Democrat Chris Murphy.
What is the Difference Between Centralized and Decentralized AI?
Centralized AI dominates today's landscape, but at the cost of privacy and censorship.
#ai #decentralized #centralized #technology
AI is developing at breakneck speed. Experts predict that between 2024 and 2030, the AI market will expand by over 28% each year. By 2030, it's expected to reach a staggering value of over $800 billion.
But this rapid progress comes with big risks. Today's AI landscape is dominated by a handful of large, centralized companies like OpenAi, Google, and Microsoft, who control most of the data and computing resources.
As the AI tools they build become more powerful, questions of privacy, security, and control are becoming urgent.
Continual data breaches at tech companies have exposed billions of user records, highlighting the vulnerabilities of centralized data storage. The Federal government, which siphons personal information from across society, is no better. centralization in AI development also leads to issues of censorship and bias, despite intention.
A new approach is emerging that tackles many of the problems with centralized AI while opening up new possibilities for innovation and user control: decentralized AI.
Centralized vs. Decentralized AI
While this approach allows for efficient processing of massive datasets, it risks data monopolies, single points of failure, and potential misuse of concentrated power.
Decentralized AI, in contrast, distributes computational resources, data, and most importantly, control, across a many participants.
This offers several key advantages:
Distributed Compute: Instead of relying on centralized data centers, decentralized AI leverages a network of distributed nodes for processing.
Data Sovereignty: Decentralized systems prevent data from collecting in single databases, under the control of singular monolithic parties. ,
Resilience: Distributed systems are inherently more resistant to single points of failure, improving overall system robustness. This resilience is achieved through redundancy and load balancing across multiple nodes, ensuring that if one part of the network fails, the system can continue operating by redistributing tasks to other available nodes.
Democratized innovation: By lowering barriers to entry, decentralized AI can stimulate a more diverse ecosystem of developers and applications, fueling innovation and greater user optionality.
transparency: Many decentralized AI initiatives emphasize open-source development, enabling greater scrutiny and community-driven improvements.
It's crucial to note that while decentralization can enhance privacy and security, these are not inherent features.
Proper implementation of privacy-preserving techniques are also necessary to fully realize the benefits, which was the entire purpose of Venice’s creation.Decentralization provides a crucial foundation that can support privacy-enhancing technologies, but privacy itself requires conscious design choices and implementation.
Here's a comparison of centralized and decentralized AI infrastructures:
This shift from centralized to decentralized infrastructure represents a fundamental reimagining of how AI systems are developed, deployed, and governed. Decentralized AI has the potential to create a more open, resilient, and democratized AI ecosystem.
Key technologies enabling Decentralized AI
Decentralized AI integrates a number of key technologies to achieve its goals of distributed processing, enhanced data security, and open innovation:
Open-Source: While not exclusively tied to decentralized AI, open-source models often play a complementary role. They enhance transparency, allow for community audits and improvements, and align with the ethos of decentralization. This is why Venice uses open-source models, but it's important to note that not all decentralized AI systems use open-source models, and not all open-source models are part of decentralized systems.
Decentralized compute: These networks allow individuals and organizations to contribute unused compute power for AI tasks, increasing overall capacity, reducing costs, and improving availability of computational resources. By turning compute power into a shared, tradable resource, it democratizes access to AI infrastructure.
Federated learning: This technique allows AI models to be trained across multiple decentralized devices or servers without exchanging raw data. Federated learning can enhance data privacy by keeping sensitive information local while still benefiting from diverse datasets, but it requires careful implementation to fully realize these privacy benefits.
edge computing: By processing data on local devices rather than in centralized cloud servers, edge computing reduces latency in AI inference and can enhance data locality. This is particularly crucial for applications requiring real-time processing or operating in environments with limited connectivity.
peer-to-peer (P2P) networks: P2P architectures enable direct interaction between network participants without the need for intermediaries. In decentralized AI, P2P networks can facilitate distributed computing, data sharing, and collaborative model training.
blockchain: blockchains provide secure and transparent ways to record transactions and share data across networks. In the context of AI, they can ensure data provenance, create auditable trails of AI decisions, and enable secure peer-to-peer sharing of AI data.
smart contracts: These self-executing contracts with the terms of the agreement directly written into code can automate transactions and enforce rules in decentralized AI networks. They can play a crucial role in managing access rights to AI models and data, as well as in implementing fair compensation mechanisms for contributions to AI development.
These technologies work in concert to form the foundation upon which decentralized AI platforms are built, enabling a new paradigm of AI that prioritizes distributed innovation and user empowerment.
The Decentralized AI landscape
The decentralized AI landscape is rapidly evolving, with numerous projects addressing different aspects of the AI stack. Here's an overview of some key players across various categories:
image credit
Decentralized Compute
Akash: Decentralized cloud computing marketplace
Morpheus: Peer-to-peer network for general purpose AI and smart agents
TogetherAI: Collaborative AI development platform
Decentralized Storage
Filecoin: Decentralized network for file storage
Arweave: Permanent, decentralized data storage
AI Infrastructure
Ocean protocol: Decentralized data exchange protocol
Oasis: Platform focusing on privacy-preserving AI computing
SingularityNET: Marketplace for AI services
Decentralized Data and Training: @Suyu0x
Gensyn: Decentralized machine learning network
Bittensor: Peer-to-peer network for AI training
Nous Research, known for developing high-quality open-source language models, is advancing distributed AI training methods. Their innovation, DisTrO (Distributed Training Over-the-Internet), is a new optimizer allowing AI models to be trained across distributed networks using consumer-grade Internet connections. At Venice, we host the latest Nous models - try them here.
These platforms are working to create a more open, transparent, and decentralized AI ecosystem. In the next section, we'll dive deeper into how Venice leverages this ecosystem and what this means for our users.
Venice's approach to Decentralized AI
Venice leverages decentralized infrastructure to provide a private, uncensored alternative to mainstream AI apps from centralized tech companies. Our platform is built on tOP of decentralized systems, enabling us to offer powerful AI capabilities while prioritizing user privacy, freedom of inquiry, and permissionless access.
Here's how we're implementing decentralized AI principles:
Decentralized inference
We utilize networks of distributed GPU providers for AI inference, such as Akash Network. When using Venice you can clearly see which provider was used to generate the output.
This approach ensures that no single entity controls our computational resources, enhancing the overall resilience and scalability of our AI services.
Cryptocurrency Payment Options
We've implemented cryptocurrency as a payment option, extending our commitment to decentralization to the financial aspect of our services. This approach supports a decentralized financial ecosystem and enables increased permissionless access to AI services globally.
By integrating cryptocurrency payments, we're not only offering an alternative payment method, but ensuring our financial infrastructure aligns with our core principle of open, decentralized AI access.
Privacy-first infrastructure
Our entire system is architected with privacy as a foundational principle. All user conversations and data remain in your local browser, never touching our servers. We don't see or store user prompts or AI responses. This design gives users full control over their AI interactions, significantly reduces the risk of data breaches or unauthorized access, and enables users to explore AI capabilities without compromising their privacy.
By building on decentralized systems, Venice offers a platform where users can access powerful AI capabilities while maintaining control over their privacy and freedom of exploration. This approach enhances individual user experiences but also contributes to a more open, diverse, and resilient AI ecosystem.
The future of AI is decentralized
Decentralized AI represents a fundamental shift in how we approach artificial intelligence, offering a compelling alternative to the centralized models that currently dominate the landscape.
Venice leverages decentralized AI systems to deliver powerful AI capabilities while prioritizing user privacy, freedom of inquiry, and uncensored access. By distributing computational resources, implementing privacy-first infrastructure, and supporting cryptocurrency payments, we're demonstrating that it's possible to create an AI platform that contributes to a more open, permissionless and decentralized AI ecosystem.
The journey towards fully realizing the potential of decentralized AI is just beginning, with many challenges in scalability, interoperability, and regulation still to be addressed. However, we believe that continued collaboration across the decentralized AI ecosystem will overcome these hurdles.
The future of AI is decentralized and permissionless, and Venice will continue building with this ethos.
Meta Introduces Spirit LM open source model that combines text and speech inputs/outputs
Just in time for Halloween 2024, Meta has unveiled Meta Spirit LM, the company’s first open-source multimodal language model capable of seamlessly integrating text and speech inputs and outputs.
#meta #spirit #lm #text #speech #technology
What is Meta Spirit LM?
Meta Spirit LM is a multimodal AI model developed by Meta's Fundamental AI Research (FAIR) team. It is designed to generate more expressive and natural-sounding speech, while also performing cross-modal tasks like speech-to-text and text-to-speech.
How does it work?
Traditional AI models for voice rely on automatic speech recognition (ASR) to process spoken input before synthesizing it with a language model, which is then converted into speech using text-to-speech (TTS) techniques. However, this process can result in speech that lacks the nuances of human communication, such as tone and emotion.
Meta Spirit LM introduces a more advanced solution by incorporating phonetic, pitch, and tone tokens, which enable the model to capture the complexities of human speech and reflect them in its generated speech. The model is trained on a combination of text and speech datasets, allowing it to perform cross-modal tasks like speech-to-text and text-to-speech while maintaining the natural expressiveness of speech in its outputs.
Phonetic, Pitch, and Tone Tokens
Phonetic tokens are used to process and generate speech by representing the sounds of spoken language as numerical vectors. Pitch and tone tokens are used to capture the nuances of human speech, such as tone and emotion, by representing the pitch and tone of spoken language as numerical vectors.
The Spirit LM model incorporates these tokens to capture the complexities of human speech and reflect them in its generated speech. The model is trained on a combination of text and speech datasets, allowing it to learn the relationships between phonetic, pitch, and tone tokens and generate speech that is more expressive and natural-sounding.
Spirit LM Base and Spirit LM Expressive
Meta has released two versions of Spirit LM: Spirit LM Base and Spirit LM Expressive. The Base version uses phonetic tokens to process and generate speech, while the Expressive version includes additional tokens for pitch and tone, allowing the model to capture more nuanced emotional states and reflect them in its output.
Spirit LM Base is a more basic version of the model that uses phonetic tokens to process and generate speech. It is designed to be more efficient and scalable than the Expressive version, while still providing a high level of expressiveness and natural-sounding speech.
Spirit LM Expressive, on the other hand, is a more advanced version of the model that includes additional tokens for pitch and tone. This allows the model to capture more nuanced emotional states, such as excitement or sadness, and reflect them in its output.
Cross-Modal Tasks
Meta Spirit LM is designed to perform cross-modal tasks, including speech-to-text and text-to-speech. Speech-to-text involves converting spoken language into written text, while text-to-speech involves generating spoken language from written text.
The Spirit LM model is trained on a combination of text and speech datasets, allowing it to learn the relationships between phonetic, pitch, and tone tokens and generate speech that is more expressive and natural-sounding. This enables the model to perform cross-modal tasks with high accuracy and natural-sounding speech.
Open-Source and Non-Commercial Availability
Meta Spirit LM is fully open-source, providing researchers and developers with the model weights, code, and supporting documentation to build upon. The model is available for non-commercial usage under Meta's FAIR Noncommercial Research License, which grants users the right to use, reproduce, modify, and create derivative works of the Meta Spirit LM models, but only for non-commercial purposes.
This open-source availability is designed to encourage the broader research community to explore new possibilities for multimodal AI applications, while also providing a high level of flexibility and customization for researchers and developers.
Applications and Future Potential
Meta Spirit LM has significant implications for various applications, including virtual assistants, customer service bots, and other interactive AI systems where more nuanced and expressive communication is essential.
The model's ability to detect and reflect emotional states like anger, surprise, or joy in its output makes it an ideal solution for creating more human-like and engaging AI interactions. This is particularly relevant in applications such as customer service, where the ability to recognize and respond to emotional cues can improve the overall user experience.
In addition to its applications in human-computer interaction, Meta Spirit LM also has potential in areas such as speech therapy, where it could be used to help individuals with speech disorders or language impairments.
Broader Effort
Meta Spirit LM is part of a broader set of research tools and models that Meta FAIR is releasing to the public. This includes an update to Meta's Segment Anything Model 2.1 (SAM 2.1) for image and video segmentation, which has been used across disciplines like medical imaging and meteorology, and research on enhancing the efficiency of large language models.
This broader effort is designed to advance the state of the art in AI research, while also providing a high level of flexibility and customization for researchers and developers. By releasing these models and tools under open-source licenses, Meta is encouraging the broader research community to explore new possibilities for multimodal AI applications.
Article
Why AI is going nuclear
If you’re of a certain age, the words “nuclear energy” probably conjure up dystopian images of power plants melting down, glowing radioactive waste, protesters, and other dark scenes ranging from the unfortunate to apocalyptic.
#ai #nuclear #energy #technology
The truth is, nuclear power’s reputation has been mostly unfairly blemished since 1970s and ’80s thanks to the Three Mile Island and Chernobyl meltdowns in Pennsylvania and Ukraine (at that time, part of the Soviet Union), respectively. While terrible, these disasters belie nuclear energy’s true safety record, which is actually much better for humans and of course, the environment, than most other power sources — even renewables, and even accounting for the fact that nuclear waste needs to go somewhere.
Now in the year 2024, some of the largest technology companies on Earth are ready to embrace nuclear power again — and the reason is because of artificial intelligence (AI).
Looking over the last 9-10 months, and in particular, the last few weeks, Microsoft, Google, and Amazon have all announced large-scale commitments to buy, invest in, and/or help build new nuclear power plants. It’s no coincidence these rivals are the three top providers of cloud computing and cloud storage solutions in the world, and have also been among the biggest to embrace and provide AI models and technology to customers, both other businesses and end-users.
Article
Passionfroot is a marketplace for business-focused content creators looking for brand partnerships — and vice versa
As the creator economy is growing rapidly, brand partnerships remain one of the prime ways creators can earn money.
As the creator economy is growing rapidly, brand partnerships remain one of the prime ways creators can earn money. Other services like link-in-bio apps with affiliate links or Pateron-like subscriptions become secondary ways to increase creators’ income. For platforms and startup, the biggest challenge remains to match brands and creators for collaboration.
#contentcreator #passionfroot #technology #economy
Berlin-based startup Passionfroot is building a toolkit and a marketplace for business, productivity, and thought leadership-focused creators for brand collaborations.
The company was founded by CEO Jen Phan, who had a career in early-stage investing and Lorenzo De Nobili, who previously started a marketplace connecting users with different professionals for services. The startup has raised $3.8 million in a seed funding round led by VC Supernode Global, with other investors including Miro CEO Andrey Khusid’s s16vc, Sequoia, Accel’s Scout Fund, and Creandum also participating — Creandum previously invested in the company’s $3.4 million pre-seed round too. The seed round included angels such as former Zapier CMO Kieran Flanagan, ex-community and creator lead at Notion Ben Lang, Linktree CPO Jiaona Zhang, and Austin Lau, who looks after Growth at Anthropic.
Before launching Passionfroot, Phan launched a tech newsletter in 2020 for tech professionals with immigration backgrounds. She thought about leaving her VC role and going full-time on the newsletter.
Article
Meta tests facial recognition for spotting 'celeb-bait' ads scams and easier account recovery
Meta is expanding tests of facial recognition as an anti-scam measure to combat celebrity scam ads and more broadly, the Facebook owner announced Monday.
Meta is expanding tests of facial recognition as an anti-scam measure to combat celebrity scam ads and more broadly, the Facebook owner announced Monday.
#meta #facial #fraud #deepfake #technology
Monika Bickert, Meta’s VP of content policy, wrote in a blog post that some of the tests aim to bolster its existing anti-scam measures, such as the automated scans (using machine learning classifiers) run as part of its ad review system, to make it harder for fraudsters to fly under its radar and dupe Facebook and Instagram users to click on bogus ads.
“Scammers often try to use images of public figures, such as content creators or celebrities, to bait people into engaging with ads that lead to scam websites where they are asked to share personal information or send money. This scheme, commonly called ‘celeb-bait,’ violates our policies and is bad for people that use our products,” she wrote.
“Of course, celebrities are featured in many legitimate ads. But because celeb-bait ads are often designed to look real, they’re not always easy to detect.”
The tests appear to be using facial recognition as a back-stop for checking ads flags as suspect by existing Meta systems when they contain the image of a public figure at risk of so-called “celeb-bait.”
“[W]e will try to use facial recognition technology to compare faces in the ad against the public figure’s Facebook and Instagram profile pictures,” Bickert wrote. “If we confirm a match and that the ad is a scam, we’ll block it.”
Article
Stability claims its newest Stable Diffusion models generate more 'diverse' images
AI startup Stability AI has announced its newest family of image generation models, Stable Diffusion 3.5, with several promised improvements.
Following a string of controversies stemming from technical hiccups and licensing changes, AI startup Stability AI has announced its latest family of image generation models.
#stability #ai #technology #imagegeneration
While Stable Diffusion 3.5 Large and 3.5 Large Turbo are available today, 3.5 Medium won’t be released until October 29.
Stability says that the Stable Diffusion 3.5 models should generate more “diverse” outputs — that is to say, images depicting people with different skin tones and features — without the need for “extensive” prompting.
“During training, each image is captioned with multiple versions of prompts, with shorter prompts prioritized,” Hanno Basse, Stability’s chief technology officer, told TechCrunch in an interview. “This ensures a broader and more diverse distribution of image concepts for any given text description. Like most generative AI companies, we train on a wide variety of data, including filtered publicly available datasets and synthetic data.”
Article
The new Stable Diffusion 3.5 series is more customizable and versatile than Stability’s previous-generation tech, the company claims — as well as more performant. There are three models in total:
#stablediffusion #technology #imagegeneration
ChatGPT’s Canvas now shows tracked changes
ChatGPT’s Canvas feature allows users to edit the chatbot’s responses on the app rather than copying and pasting them to a separate document.
However, when Canvas launched in early October for its paid tiers, it didn’t let people see what changes GPT-4o made to its responses. OpenAI’s latest update to the feature corrects that.
#chatgpt #canvas #technology #ai #openai
The show changes button will show the most recent changes to either the generated text or code on Canvas. It will highlight added information in green and deleted sections in red.
Tracking changes has always been a good feature of any editing platform; Google Docs and Word documents offer a toggle for users to check what’s been changed. But OpenAI had been planning to roll out updates to Canvas slowly as ChatGPT subscribers get used to it.
Canvas already offers familiar features like comments, where users can add suggestions or give more instructions for the AI model to follow when editing responses.
Canvas is still only available on the web version of ChatGPT for ChatGPT Plus, Teams, Enterprise and Edu users. Mac app users and anyone downloading the recently released Windows version of ChatGPT will have to wait until Canvas is rolled out to these standalone apps.
Currently, people can access Canvas on the regular ChatGPT window rather than in any custom GPTs.
Article
Anthropic Wants Its AI Agent to Control Your Computer
Claude is the first major AI model to be able to take control of a computer to do useful work.
Demos of AI agents can seem stunning but getting the technology to perform reliably and without annoying, or costly, errors in real life can be a challenge. Current models can answer questions and converse with almost human-like skill and are the backbone of chatbots such as OpenAI’s ChatGPT and Google’s Gemini. They can also perform tasks on computers when given a simple command by accessing the computer screen as well as input devices like a keyboard and trackpad or through low-level software interfaces.
#aiagents #anthropic #claude #technology
Anthropic says that Claude outperforms other AI agents on several key benchmarks including SWE-bench, which measures an agent's software development skills and OSWorld, which gauges an agent's capacity to use a computer operating system. The claims have yet to be independently verified. Anthropic says Claude performs tasks in OSWorld correctly 14.9 percent of the time. This is well below humans, who generally score around 75 percent, but considerably higher than the current best agents, including OpenAI’s GPT-4, which succeed roughly 7.7 percent of the time.
Anthropic claims that several companies are already testing the agentic version of Claude. This includes Canva, which is using it to automate design and editing tasks and Replit, which uses the model for coding chores. Other early users include The Browser Company, Asana and Notion.
Ofir Press, a postdoctoral researcher at Princeton University who helped develop SWE-bench, says that agentic AI tends to lack the ability to plan far ahead and often struggle to recover from errors. “In order to show them to be useful we must obtain strong performance on tough and realistic benchmarks,” he says, like reliably planning a wide range of trips for a user and booking all the necessary tickets.
Article
A Lawsuit Against Perplexity Calls Out Fake News Hallucinations
In a new copyright lawsuit against AI startup Perplexity, Dow Jones and the New York Post argue that hallucinating fake news and attributing it to real papers is illegal.
#perplexity #ai #technology
In a statement emailed to WIRED, News Corp chief executive Robert Thomson compared Perplexity unfavorably to OpenAI. “We applaud principled companies like OpenAI, which understands that integrity and creativity are essential if we are to realize the potential of Artificial Intelligence,” the statement says. “Perplexity is not the only AI company abusing intellectual property and it is not the only AI company that we will pursue with vigor and rigor. We have made clear that we would rather woo than sue, but, for the sake of our journalists, our writers and our company, we must challenge the content kleptocracy.”
OpenAI is facing its own accusations of trademark dilution, though. In New York Times v. OpenAI, the Times alleges that ChatGPT and Bing Chat will attribute made-up quotes to the Times, and accuses OpenAI and Microsoft of damaging its reputation through trademark dilution. In one example cited in the lawsuit, the Times alleges that Bing Chat claimed that the Times called red wine (in moderation) a “heart-healthy” food, when in fact it did not; the Times argues that its actual reporting has debunked claims about the healthfulness of moderate drinking.
Article
SAP boss warns against regulating AI, says Europe risks falling behind U.S., China
Christian Klein, head of German software giant SAP, says Europe risks falling behind the U.S. and China if it ends up overregulating the AI sector.
Europe should avoid regulating artificial intelligence and focus its attention on the results of the technology instead, the CEO of German enterprise tech giant SAP told CNBC Tuesday.
#sap #technology #europe
Christian Klein, who has held the top job at SAP since April 2020, said Europe risks falling behind the U.S. and China if it overregulates the AI sector.
While it's important to mitigate the risks associated with AI, Klein argued that regulating the tech while it's still in its infancy would be misguided.
"It's very important that how we train our algorithms, the AI use cases we embed into the businesses of our customers — they need to deliver the right outcome for the employees, for the society," Klein said on CNBC's "Squawk Box Europe" Tuesday.
"If you only regulate technology in Europe, how can our startups here in Europe, how can they compete against the other startups in China, in Asia, in the U.S.?" Klein added.
"Especially for the startup scene here in Europe, it's very important to think about the outcome of the technology but not to regulate the AI technology itself."
Article
More startups are being spun out of Klarna than any other European fintech unicorn
Alumni from Klarna have gone on to create 62 new startups — more than any other fintech unicorn in Europe, according to a new report from VC firm Accel.
More startups are being spun out of Swedish digital payments firm Klarna than any other financial technology unicorn in Europe, according to a new report from venture capital firm Accel.
#fintech #klarna #europe #technology
Accel's "Fintech Founder Factory" report shows that alumni from Klarna have gone on to create a total of 62 new startups, including the likes of Swedish lending technology firm Anyfin, regulatory compliance platform Bits Technology and AI-powered coding platform Pretzel AI.
That is more than any other venture-backed fintech startup worth $1 billion or more in the region.
This includes the digital banking app Revolut, whose former employees have founded 49 startups. It also includes money transfer app Wise and online-only bank N26, where ex-staff at both firms have started 33 companies each, according to Accel's data.
Article
U.S. proposals to curb AI, semiconductor investment into China under final review
The proposals are part of the Joe Biden administration's efforts to restrict the flow of U.S. capital, technology and expertise into China.
Measures aimed at curbing U.S. investments into China in sensitive technologies are in the final stage of review, a U.S. government update showed.
#china #semiconductor #technology #unitedstates
Under this set of rules, the Treasury Department will require notification of outbound investments into China in sensitive technologies including artificial intelligence, semiconductors, microelectronics and quantum computing that can be employed for developing military capabilities.
The final rules will likely be released within the "next week or so," according to Reuters.
The proposals are part of the Joe Biden administration's efforts to restrict the flow of U.S. capital, technology and expertise into China that could support its military modernization and undermine U.S. national security.
In June last year, U.S. Treasury Department released proposals that include potential outright bans on certain investments into China in these cutting-edge technologies.
"The potential military, intelligence, surveillance, and cyber-enabled applications of these technologies and products pose risks to U.S. national security particularly when developed by a country of concern such as the PRC," the Treasury Department notification said.
Former Treasury official Laura Black said the department could be trying to make the rules official before the presidential election — which is set to take place on Nov. 5 — Reuters reported.
Article
GE HealthCare announces time-saving AI tool for doctors who treat cancer
GE HealthCare announced a new artificial intelligence application it said will save time for doctors who diagnose and treat cancer.
GE HealthCare on Monday announced a new artificial intelligence application it said will save time for doctors who diagnose and treat cancer.
#healthcare #ge #ai #technology
CareIntellect for Oncology, as the tool is called, will help oncologists get up to speed on a patient's history and disease progression by quickly showing them the data they need, the company said. GE HealthCare wants to spare oncologists the headache of digging through records so they can focus on caring for their patients, the company said.
Health-care data is notoriously difficult to analyze, and as much as 97% of the data produced by hospitals goes unused, according to a Deloitte report. That information is stored across numerous vendors and file formats such as images, lab test results, clinical notes and device readings, which can be extremely taxing for doctors to sort through.
"It's very time-consuming, very frustrating for these clinicians," Dr. Taha Kass-Hout, GE HealthCare's global chief science and technology officer, told CNBC in an interview.
CareIntellect for Oncology will be able to summarize clinical reports and identify when patients are deviating from their treatment plans, Kass-Hout said. The system can flag when a patient misses a lab test, for instance, so that their doctor can determine the best next steps.
Article
Microsoft to roll out new autonomous AI agents next month, fending off challenge from Salesforce
Microsoft will allow organizations to create their own autonomous AI agents within Copilot starting next month.
Microsoft will allow businesses to start making their own autonomous artificial intelligence agents starting next month, taking the fight back to Salesforce, which introduced its own configurable agentic AI tools in September.
#microsoft #aiagents #copilot
At its "AI Tour" event in London on Monday, Microsoft revealed plans to allow organizations to create their own autonomous agents within Copilot Studio, the U.S. tech giant's platform for customizing and building so-called "copilot" assistants.
These agents had previously been available in private preview after Microsoft announced them initially in May. Starting next month, they'll move into public preview, meaning more organizations can start building AI agents of their own.
AI agents can act as virtual workers that can carry out a series of tasks without supervision. They are touted as a major evolution of large language model-based AI from chat interfaces, creating an experience that blends more seamlessly into the background.
Beyond adding the ability to create autonomous agents in Copilot Studio, Microsoft said it would also launch 10 new autonomous agents in Dynamics 365, the company's suite of enterprise resource planning and customer relationship management apps.
Article
Interface.ai raises $30M to help banks field customer requests
Interface, a startup building an AI-powered platform to help banks field customer service requests, has raised $30 million in a funding round.
Interface.ai, a customer automation platform for banks and other financial institutions, today announced that it closed a $30 million funding round led by Avataar Venture Partners.
#interfaceai #ai #banks #technology
$20 million of the round was equity, while the remaining $10 million was in the form of debt. It’s Interface’s first outside capital; the startup had been completely bootstrapped.
“Interface is trusted by more than 100 financial institutions across North America, processing millions of interactions every day,” CEO and co-founder Srinivas Njay told TechCrunch. “We’re generating tens of millions in annual recurring revenue.”
Njay said that the inspiration for Interface came from his father, who owned a credit union in India. Previously, Njay worked at Microsoft as a product manager on the Bing advertising team and at EA as a senior product manager on mobile games.
Njay teamed up with Bruce Kim, the founder of customer management and billing company Inovaware, in 2019 to launch Interface.
Article
WhatsApp introduces contact storing directly within the app, teases usernames
WhatsApp on Tuesday announced a new feature that will allow users to save contacts within the app.
WhatsApp on Tuesday announced a new feature that will allow users to save contacts within the app. That means that even if you lose your phone or link a new device to your primary number, you will see all your contacts stored within WhatsApp’s cloud storage.
#whatsapp #application #technology #meta
Until now, WhatsApp relied on your phone’s contact book to sync contacts. If a new person messaged you on the app, and you wanted to save their details, the information would be stored locally on your device.
The new feature, rolling out soon, lets you store contacts on any device, including WhatsApp for the web and Windows. You can choose to sync contacts saved within WhatsApp to your phone, as well.
Last year, the company began allowing users to log into two WhatsApp accounts on one device. Saving contacts linked to one WhatsApp account is handy if you want to keep personal and business contacts separate. Plus, if you share the device with others, you can maintain your own contact list linked to your number.
Article
Small but mighty: H2O.ai’s new AI models challenge tech giants in document analysis
H2O.ai, a provider of open-source AI platforms, announced today two new vision-language models designed to improve document analysis and optical character recognition (OCR) tasks.
The models, named H2OVL Mississippi-2B and H2OVL-Mississippi-0.8B, show competitive performance against much larger models from major tech companies, potentially offering a more efficient solution for businesses dealing with document-heavy workflows.
#h2o #ai #technology #h2oai
The H2OVL Mississippi-0.8B model, with only 800 million parameters, surpassed all other models, including those with billions more parameters, on the OCRBench Text Recognition task. Meanwhile, the 2-billion parameter H2OVL Mississippi-2B model demonstrated strong general performance across a range of vision-language benchmarks.
“We’ve designed H2OVL Mississippi models to be a high-performance yet cost-effective solution, bringing AI-powered OCR, visual understanding, and Document AI to businesses,” Sri Ambati, CEO and Founder of H2O.ai said in an exclusive interview with VentureBeat. “By combining advanced multimodal AI with efficiency, H2OVL Mississippi delivers precise, scalable Document AI solutions across a range of industries.”
The release of these models marks a significant step in H2O.ai’s strategy to make AI technology more accessible. By making the models freely available on Hugging Face, a popular platform for sharing machine learning models, H2O.ai is allowing developers and businesses to modify and adapt the models for specific document AI needs.
Article
Qualcomm touts Snapdragon 8 Elite chip's massive performance gains and AI capabilities
Qualcomm has introduced the Snapdragon 8 Elite at its annual Snapdragon Summit, touting it as the world's fastest mobile CPU that delivers unprecedented performance and efficiency.
s the mobile industry continues to evolve, the Snapdragon 8 Elite represents a major step forward in processing power and efficiency. Its impact on the next generation of smartphones and mobile devices will undoubtedly be significant, potentially reshaping user experiences and expectations.
#qualcomm #snapdragon #chip #technology
Qualcomm has introduced the Snapdragon 8 Elite at its annual Snapdragon Summit, touting it as the world's fastest mobile CPU that delivers unprecedented performance and efficiency.
The chipset showcases Qualcomm's second-generation custom Oryon CPU, engineered to handle complex multimodal AI tasks. It boasts Oryon CPU Prime cores running at 4.32GHz, complemented by six performance cores reaching speeds of 3.53GHz. A 24MB L2 cache further enhances its processing power. The Hexagon NPU delivers 12 times the performance of its predecessor, while the Adreno GPU offers a threefold improvement in performance.
Qualcomm says it has optimized the Snapdragon 8 Elite for various use cases, particularly in gaming, AI processing, and photography.
For gaming enthusiasts, the chip promises smoother gameplay with higher framerates while extending gaming time by up to 2.5 hours, thanks to a 40 percent improvement in both battery efficiency and GPU performance.
In terms of AI processing, the new Qualcomm AI Engine leverages the Oryon CPU for latency-critical AI tasks, while the Hexagon NPU provides faster inferencing performance.
Article
99% of AI companies won't survive bubble burst
Baidu CEO Robin Li believes that only about one percent of AI companies will endure after the inevitable burst of the AI bubble.
The rising popularity of generative AI has given place to the development of numerous startups making bold claims about the technology's capabilities. Many onlookers liken the trend to a classic market bubble, a sentiment shared by the CEO of Chinese tech giant Baidu. However, when the lofty expectations for AI collide with reality, most AI startups may not survive.
#ai #baidu #bubble #technology
The rising popularity of generative AI has given place to the development of numerous startups making bold claims about the technology's capabilities. Many onlookers liken the trend to a classic market bubble, a sentiment shared by the CEO of Chinese tech giant Baidu. However, when the lofty expectations for AI collide with reality, most AI startups may not survive.
Baidu CEO Robin Li believes that only about one percent of AI companies will endure after the inevitable burst of the AI bubble. Li suggests this will result in a healthier market with more realistic applications for the emerging technology.
Generative AI has remained a hot topic in the years following its initial surge, driven by companies like OpenAi and NVIDIA. The latter, which produces the hardware that most AI companies depend on, could soon become the world's most valuable company. Tech giants such as Microsoft, Apple, and Google are also integrating AI into nearly all of their future products.
However, many companies often make ambitious claims about the jobs AI might replace or what users can create using text prompts.
Speaking at Harvard's Future of business Conference, Li argued that many of these products will turn out to be false innovations, unable to find a sustainable market.
He compared the current situation to the dot-com bubble that burst around the turn of the century, wiping out many early Internet companies. Li also believes the AI market has become quieter but healthier in 2024 compared to recent years. Although some businesses and investors maintain high hopes for AI, others are showing more caution, recognizing that Nvidia's meteoric rise can't last forever.
Consumer enthusiasm for the technology remains lukewarm. Recent sales reports indicate that consumers aren't purchasing PCs with AI-focused hardware due to a specific interest in AI, but rather because the latest models from major vendors come equipped with the technology by default.
If a dramatic market correction occurs, Li anticipates that the remaining one percent of AI companies will offer highly valuable products and services. Addressing a persistent issue in generative AI, Li expressed optimism about the reduction of hallucinations in chatbots, claiming that they have become significantly more accurate over the past 18 months.
The CEO also touched on concerns about AI-related job losses, acknowledging the possibility that AI could replace human workers. Li believes a paradigm shift in employment may take place over the next 10 to 30 years.
Article
MIT engineers create solar-powered desalination system producing 5,000 liters of water daily
Engineers at the Massachusetts Institute of Technology (MIT) have created a new desalination system that harnesses the sun's power without requiring backup batteries.
The MIT engineers' solar-powered desalination system represents a significant leap forward in sustainable water treatment technology. By eliminating the need for batteries and maximizing the use of solar energy, the system can be deployed in regions lacking reliable power infrastructure.
#mit #desalination #solar #water #technology
The desalination system developed by the MIT team is a significant breakthrough in harnessing renewable energy to produce clean drinking water. Here's a more detailed explanation of the technology:
How it works
The system uses a combination of electrodialysis and solar power to remove salt ions from water. Electrodialysis is a method that uses an electric field to separate salt ions from water, a process known as ion exchange. The MIT team's system consists of three main components:
Control strategy
The system's control strategy, dubbed "flow-commanded current control," allows it to dynamically adjust its desalination rate multiple times per second. This is achieved through a sophisticated algorithm that monitors the system's energy output and adjusts the flow rate of water through the electrodialysis stacks accordingly.
When the system detects an excess of available power, it increases the rate at which water is pumped through the electrodialysis stacks, allowing for more efficient desalination. Conversely, when the system detects a lack of power, it reduces the flow rate to conserve energy.
Efficiency and scalability
The system has been designed to be highly efficient, with an average energy conversion rate of 94%. This means that 94% of the electrical energy generated by the solar panels is converted into clean drinking water.
The system's scalability is also significant, with the potential to supply large communities and even entire municipalities with low-cost, sun-powered drinking water. The MIT team has already demonstrated the system's effectiveness in a small community setting, where it supplied Enough water for a population of around 3,000 people.
Real-world testing
The MIT team has tested the system in real-world conditions, deploying a community-scale prototype at groundwater wells in New Mexico. The prototype operated for six months under variable weather conditions and with different water types, demonstrating the system's feasibility and effectiveness.
The results of the testing were promising, with the system producing high-quality drinking water that met or exceeded regulatory standards. The team also observed significant reductions in energy consumption and greenhouse gas emissions compared to traditional desalination systems.
Potential impact
The MIT desalination system has the potential to make a significant impact on communities around the world, particularly in regions where access to clean drinking water is limited. According to Jonathan Bessette, a MIT PhD student in mechanical engineering, "The majority of the population actually lives far enough from the coast, that seawater desalination could never reach them. They consequently rely heavily on groundwater, especially in remote, low-income regions. And unfortunately, this groundwater is becoming more and more saline due to climate change."
The system's ability to harness renewable energy and produce clean drinking water without relying on backup batteries makes it an attractive solution for communities that need reliable access to clean water.
Future plans
The MIT team plans to continue testing and scaling up the system, with the goal of deploying it in larger communities and even entire municipalities. They are also preparing to launch a company in the coming months, with the goal of pushing out the deployment of these systems to areas in need around the world.
Overall, the MIT desalination system represents a significant breakthrough in harnessing renewable energy to produce clean drinking water, and has the potential to make a significant impact on communities around the world.
Article
Decentraland Launches Revamped Virtual World with Enhanced Performance, Engaging Features, and Future-Ready Architecture
Decentraland, the first decentralized virtual social world, launched a powerful new desktop client today, signaling a new era for experiencing spatial environments. This beta release enhances performance, delivers a more immersive environment, and introduces new engaging features, setting the stage for future innovations and expanded creator tools.
#decentraland #virtualreality #gaming #crypto
This milestone marks the shift away from a browser-based experience, replaced by a more stable, Unity-powered desktop client and the new Creator Hub, aligning with Decentraland’s vision for an interconnected metaverse.
“Today marks a new chapter for Decentraland. As virtual worlds take on a larger role in digital life, Decentraland is committed to creating a virtual world where people, not corporations, own their digital lives. Traditional platforms view their users as products while limiting real ownership, but Decentraland flips this script, using decentralization to empower its users,” said Yemel Jardi, Executive Director of the Decentraland Foundation. “As the internet transforms into a more immersive, spatial experience, Decentraland is leading the movement to creating a space that champions social connection, creativity, and autonomy—transcending cultures and backgrounds. This next chapter is about unlocking the true potential of the open metaverse, ensuring that as the digital world grows, it remains user-driven, community-focused, and grounded in open standards.”
Article
The Army wants Microsoft to "substantially" lower the price of its $80,000 HoloLens-based goggles
Back in 2018, Microsoft began prototyping the Integrated Visual Augmentation Systems (IVAS) augmented reality glasses and was awarded a $480 million contract by the Army for 100,000 units.
The HoloLens-based goggles that Microsoft is selling to the US Army have had a long and difficult history. While the hardware and software issues of the early models appear to have been addressed in the latest version, the problem now is the price: the military branch wants the headsets to "be substantially less" than the current cost of $80,000 per unit.
#newsonleo #technology #microsoft #hololens #goggles #army
Back in 2018, Microsoft began prototyping the Integrated Visual Augmentation Systems (IVAS) augmented reality glasses and was awarded a $480 million contract by the Army for 100,000 units.
The initial agreement was expanded in March 2021, ensuring Microsoft will provide finalized production versions, parts, and support in an agreement potentially worth up to $21.9 billion across a decade. The Army ordered an initial 5,000 units, valued at $373 million, with around 121,000 to be delivered over the course of the deal.
The project has been far from smooth from the beginning. The goggles' rollout was pushed back from fiscal year 2021 to 2022, but the Army said it remained fully committed to the deal.
In April 2022, the DOD warned that the massive purchase could be a waste of taxpayer money. There was worse news for Microsoft that year when an army tester said that the headsets pose a danger to soldiers due to the light they generate while active, which could alert enemy forces to a wearer's location.
Article