I have to disagree that the lack of details is common with these type launches. After the Cybertruck, Model 3 and Model Y people were able to immediately preorder and see all the specifications of the vehicles. This is literally the first one I can think of that lacked details and preorders.
Tesla is not a full blown autonomy company. It's well known that the Optimus robots had human controllers at the event last night, and that Tesla has been mapping out the WB studios for weeks for this Cybercab demo. Nothing demo'd last night was actually autonomous.
I also don't think you can call Tesla an autonomy company until its making the majority of its revenue from autonomy. As far as I know Tesla hasn't made any money from autonomy yet.
By this logic, Meta is merely a social media company (actually advertising) since there is where the bulk of their revenues come from. This from the company that has one of the leading LLM platforms and other AI tools rolling out.
Semantics? Maybe. But I disagree with your assessment here.
?Tesla is not a full blown autonomy company. It's well known that the Optimus robots had human controllers at the event last night,
Well known by who? Where is the evidence? By those on shows like CNBC? Actually, I saw a robotics experts, a guy who started and sold two robotics company, who was at the show asked that question and he said he could tell. It is likely some of it was, but not all. The dancing, for example, was not human controlled. That was synchronized showing the bots were doing it on their own and that it was programmed in.
So what? Isnt this what Waymo does for its autonomous division. Did the vehicles drive themselves? Isnt that autonomous?
Ironic that which Waymo gets applauded for, when Tesla does it, they are someone how autonomous.
As for how much of that was due to mapping, that is up to speculation and nobody can say how the software arrived at its conclusions.
I would define Meta as an advertising company rather than an AI company. LLAMA is being developed in service of more accurately targeting customers with advertising. If Meta ends up earning significant revenue from selling LLMs then I'm happy to rethink that assessment.
Optimus being human controlled is well known by people close to Tesla. Adam Jonas from Morgan Stanley wrote about it, Ross Gerber has been discussing it on Twitter, as has MKBHD. Tesla employees told people Optimus were being human operated.
The dancing was totally preprogrammed, but that's not autonomy.
I saw today that Tesla has been mapping out the WB Studios for 3 months, and it looks like there is some credible evidence that the Cybercab doors being opened were all human controlled.
Personally I'm not even sure if Waymo will ultimately be successful, but Waymo vehicles are autonomous because they can navigate the chaotic roads without a driver. The cities they operate it are all mapped, but the autonomy and all their sensors can (so far) deal with all sorts of scenarios. The cybercab demonstration was synchronized and planned, there were no chaotic elements for the autonomy to navigate.
Let's be perfectly honest, $TSLA is overpriced as both auto stock and an AI stock. The cybercab launch was an attempt to distract people from the Model 2 being scrapped, the not good earnings call next week, and the various regulatory cases against Tesla. I am legitimately worried for anyone who holds $TSLA stock right now and would ask people to seriously think about derisking.
You could be right on this. Markets are notorious for being crappy at pricing value. Of course, you could be wrong.
As for your views on autonomy, I think you are mistaken. To claim that Tesla does not have this is incorrect.
Tesla does this. It is supervised meaning someone is the vehicle. However, when engaged the car drives itself. This is not level 5 autonomy but it is autonomous. Of course, Waymo fails the definition too since it does have operator ready to step in.
Hence, Waymo does not run without operators ready to intervene. So what is the difference?
That might be true. There was likely no FSD configured for that as of yet. They just contoured it for the Cybertruck.
Unlike the vehicles with FSD which can operate the vehicle without the human "driving": As for your view of Waymo being autonomous, try to put it in a city that it is not in.
It cant. So you are comparing a prototype geofenced for a demonstration to Waymo's business model. I know you said that you arent sold on them being successful. However, you do use them as the poster child for autonomy it seems.
As for Ross Gerber, that is the guy who kept claiming that Elon diverted the GPUs to XAI over Tesla as a sign he was focused on other things. The guy failed to mention the facility in Austin was under construction.
Just another I hate Elon guy.
The difference is the legal burden.
If a Waymo vehicle causes an accident, then Waymo the company is responsible and accountable for the legal costs of that accident.
This legal burden is why they have lots of cameras, sensors, LiDAR, geofencing, mapping and redundancies built in to limit their exposure to risk.
Supervised FSD puts the legal burden on the driver/owner of the Tesla. Tesla the company isn't legally responsible for any accidents or mishaps on the road involving Tesla vehicles. Tesla vehicles don't have LiDAR, sensors and redundancies built in for safety to limit their legal burden. Supervised FSD Critical Interventions are still way too high for Tesla to assume this risk.
I think I would be way more confident in saying that Tesla has solved autonomy once they're confident in their own product enough to assume the legal burden. This risk would be absolutely massive considering the 2M+ Tesla vehicles out on the roads.