Coca-Cola created quite a stir by releasing an online ad that was generated by AI.
From a marketing perspective, we can say this was successful. The fact that people are talking about it is the goal for any marketing piece. It is unlikely that the ad will sell one more can of Coca-Cola. At the same time, we also will not see people turning away from the products because of it.
There appear to be some strong feelings on both sides. Some were impressed by the ad while others felt this was a "crime against humanity". Either way, people are going to have to realize what we are dealing with.
This is the future. Anyone who thinks this is the last venture into this realm is sadly mistaken.
It is also a indicator of where things are going with regards to video production.
Before getting started, here is the advertisement.
The Future of Video Creation is Arriving
Today is the worst that AI will ever be.
It is worth remembering this concept whenever you see something that looking fairly lame.
With this advertisement, was it impressive? To me, I would rank it as "meh". The value is not in the quality itself but in what it took to make. This was likely a fraction of the personnel (and cost) compared to a regular ad.
Some will maintain that the added cost shows how people are needed. The fact the quality lags is an indicator of technology's limitations.
Go back to the line above: today is the worst that AI will ever be.
Anyone who follows the progress of generative AI realizes this is nothing more than a phase we are going through. Compared to a year ago, this is quality stuff. Does it match up to what can be done with a team of humans? Not at the moment.
Another key is the fact that this video will be much better in 6 months. Actually, if we go out a year, I am going to forecast that a significant portion of online ads will be done using AI by next holiday season.
Coca-Cola will not be the outlier. In fact, we will look back and realize it was simply ahead of the curve.
Image generated by Ideogram
Working Up The Scale
Technology is a bottom up approach. What does this mean?
To start, we deal with low hanging fruit. When it comes to generative AI, this is text. Images, audio, and video are more difficult media to deal with. For that reason, the advancements in chatbots outpaced the other areas.
Over the past year, we did see enormous progress with images. Are they perfect? Absolutely not. However, they do suffice in certain instances; which brings us to our second point.
The focus begins where expectations are lower. The image I generated above is not of the quality that you would want to see in a museum. Actually, it doesn't even make it for a basic photography showing.
That said, it is of good enough quality to be posted in an online article. Here we come across another key.
The Internet has a different level of expectation as compared to other areas. For example, when it comes to video, cinema level quality is not required. How many hours per week does the average person spend watching YouTube, videos that will not rival major studios?
For this reason, I feel safe in making the prediction above. Online advertising does not require the same level of quality. Will advertisements for the 2026 Super Bowl be done exclusively using AI? Probably not. However, to place an ad on X or Facebook, that will likely be commonplace.
Entertainment Disruption
So far we focused upon advertising. This is just the tip of the iceberg.
The honeypot is the millions of hours of video content that is created for entertainment (or information) purposes. Here is where AI is going to dominate.
Computers speed things up. That is an advantage they will always have. It is why the discussion about things stalling (in any digital) area are usually unfounded. The fallback is always brute force, i.e. more processing.
For more than half a century, humans were second class when it came to the calculations game. Google Docs can add up a column of numbers in milliseconds whereas, depending upon the length, it could take a human a few minutes (with a calculator).
As such, we can look at computers as time compressors. What once took minutes, especially if done manually, is now near instant.
This is the path video generation will take. It is in the embryotic stage today. Nevertheless, it keeps working up the scale. We will see the advertising videos getting better. This means the quality is there, allowing for the ability to generate some low-end entertainment or informational videos.
Speed is going to be the key.
We will arrive upon a time when a feature length film will be generated in a few seconds. This is likely a few years down the road but we will see it. My guess is to watch the progression with text and images. They will improve in lockstep although to varying degrees.
The fact a major corporation is putting out an online add that is AI generated shows disruption is here. It is only going to accelerate into the future.
Remember, today is the worst AI will ever be.
Posted Using InLeo Alpha
interesting.
i agree it is not an outstanding commercial, but it was short, to the point, and portrayed Coca COla in a favorable an neutral fashion.
Th real buzz was that this is AI now, so what will these type of Advs look like in the future. ?!
If you prompt the AI to make incremental changes in a serial fashion it can keep making versions of this video, sort of a endless loop of reiteration burning up your resource credits on the AI platform.
But each time, each reieration of this video will seeking to improve some factor, which you ask AI to concentrate on, like views, or engagement with the video.
The future of video AI products is bright.
We all need to get on this train, called AI, before it leaves the station withpt us, and there are many ways to do just that...
....Including asking AI... LOL
That is true. Reprompting things and targeting improvements certainly is a pathway there. We see this with longer videos, i.e. movies, that use CGI (and not generative AI). The entire production is no done with a computer, only certain parts.
My view is the applications will only get better, meaning the advertisement could be generated in one or two prompts. My guess is we see that by the end of next year for short videos.
Indeed I think your right, the AI Ads will only get better.
I wonder if they can make short ads, can they make short video stories?
I would say if the technology is there for short video ads, then it can be used for short stories.
My guess is we will see things heating up by the middle of next year with video generation.
I agree.
I see AI generating video simulations of celebrity Actors like Darryl from Walking Dead in roles in video games with speech from the celebrity synthesized by the AI.
If they can do this for short videos for video games, I think they can learn to piece together a series of shots to make a movie. Similar to how a director pieces together the hundreds of scne shots to make a movie.
I look forward to the middle of next year!
People find it hard to believe but, in 5 years, our movie stars will be AI generated.
I agree.
Actors and Actresses can be challenging to deal with, due to personalities and politics.
They also have to maintain ideal weight , hair, skin and other heath attributes to be the best version of themselves for roles.
Using Ai generated versions of these stars guarantees they will be the right weight, right hair, right skin and right attitude to complete the project on time, and allow occaisional cameo performances by the actors or stunt people. Done correctly it could elminate the need for stunt men or other ancillary staff.
The future is going to be different from the past.
Would you also say that the future is censorship and fragmentation?
To be more specific, one Hive platform censors replies published from the other ones. Is it the future, or the past?
Applications are free to show what they want. The key is the data is open and censorship resistant. It can always be read on a block explorer in addition to many other applications.
One can always pull the data since it is not controlled by one company or party. In fact, nobody owns it.
So is it advisable to use and promote applications that do not hide content on Hive then? And avoid the censoring one?
Certain apps are going to show certain things. So it is up to the developers to decide what they show. You use the term censor when it isnt really censoring. That is removing the content.
If someone opts not to show short form, that is the choice of the app. After all, Spliterlands is a front end that doesnt show blog posts yet it does pull hive data.
Hiveblog doesnt have short form shown either.
Hive.blog allows everybody access any short form and interact with it. Making content unaccesssible is censorship though.
Edit: When steemit.com was deciding what not to show, we called that censorship too, didn't we?
Yes and it was incorrect. The information was not censored since it was on the blockchain. That is censorship resistant. People claimed it was censored when it was not. A block explorer reveals all data is there.
As for you point here, there is nothing that says an application has to show everything. You could set up an app that only shows blog posts without comments. You could opt not to tie into the image servers and show all content without picture or video.
That's true.
We should not support apps that choose to hide content though. Just as we chose not to support steemit.com.
I think there was nothing bad about the AI advertisement by Coca Cola and agree with OP that in the future the quality standards of such AI advertisements will further be boosted.
I agree in principle, but it's more complex. Not everyone can do an ad like this for Coca-Cola. Intelligent advertising agencies have already started using this technology. It's true, however, that smaller entrepreneurs will also begin using it.
It's not a problem now, but it will probably be when convergence happens and the masses start using it in all areas of human life...
I'm also interested to see what Musk will do with the D.O.G.E. agency; I'm sure he's got the AI (Grok?) behind him when he announces the sacking of over a million civil servants. If he succeeds there, it will succeed in all private business. If it doesn't, it could be total chaos and blood.
I am not sure Grok has much to do with his position with DOGE. They are completely separate entities.
Plus DOGE is not a true department. It is an advisory role. Musk has no power to implement anything with the government.
I don't know anything about the GROK position, but I'm pretty sure he will use some kind of AI for that optimization.
Tottaly agree.
Every point you made is spot on.
Ofcourse ai is going to dominate almost every industry sooner or later...
I completely agree. Of course, Coca-Cola doesn't sell drinks by advertising. Everyone knows who they are. For them, it's about brand awareness and top-ot-mind. Keep their name front and center is all that it's about. AI-generated ads reduces costs, which increases profits for shareholders. For Mom & Pop, it could be a lifesaver.
This post has been manually curated by @bhattg from Indiaunited community. Join us on our Discord Server.
Do you know that you can earn a passive income by delegating your Leo power to @india-leo account? We share 100 % of the curation rewards with the delegators.
100% of the rewards from this comment goes to the curator for their manual curation efforts. Please encourage the curator @bhattg by upvoting this comment and support the community by voting the posts made by @indiaunited.
It’s already here. Many of us are earning $1000 a day with your Ai generated YouTube Shorts on YouTube. It’s incredible. And lots of fun.
What??