[peutetre]

> Previously, customers were offered the option to purchase “Full Self-Driving Capability,”

They were offered a lie. Tesla still hasn't delivered what those customers paid for.

> Let’s look at what else has changed on Tesla’s website on FSD before we dive into the wording changes.

Another recent change on Tesla's website is to remove old blog posts, including a 2016 blog post in which Tesla claimed "as of today, all Tesla vehicles produced in our factory – including Model 3 – will have the hardware needed for full self-driving capability at a safety level substantially greater than that of a human driver":

https://electrek.co/2024/08/24/tesla-deletes-its-blog-post-s...

https://web.archive.org/web/20240709163806/https://www.tesla...

Tesla might now also outsource its AI work to xAI:

https://www.wsj.com/tech/tesla-xai-partnership-elon-musk-30e...

If indeed Tesla is "worth basically zero" without full self-driving:

https://electrek.co/2022/06/15/elon-musk-solving-self-drivin...

Then moving that work to xAI seems like a good way to turn Tesla into a private company without actually purchasing it.

[davej]

> Another recent change on Tesla's website is to remove old blog posts, including a 2016 blog post in which Tesla claimed …

Perhaps unintended but this is a bit misleading. Tesla changed their blog system and didn’t migrate older posts. My initial reading of your comment was that they selectively removed some older posts which they wanted to hide.

[bravetraveler]

My reading of your post is they changed systems for plausible deniability

[piva00]

Even more since migrating content from one CMS to another is a trivial engineering effort.

[bravetraveler]

Considering they decided it was worth upgrading, yeah. A marginal amount of time to backup, normalize, and import the data isn't a lot.

Not worth keeping the posts? No readership? Why upgrade?

Sus, as the kids say, but plausible.

[daghamm]

Some Teslas don't have the sensor hardware and the compute power to do road-safe FSD. This is something Tesla engineers learned the hard way.

Things may change in the future as we make advances in computing and AI, but right now it is not possible.

[mrich]

So it's ok to sell a feature to customers if you only find out later it's not possible?

[jvanderbot]

A more generous interpretation is that GP provides context and logical reasons behind the decision, not excuses for it.

[epolanski]

It's Elon Musk after all.

His boldness and thinking he's exceptional is both the reason of his success and failures.

[hdjjhhvvhga]

To be fair I believe we all have this basic bias. For example, when I have a streak of failures, I approach another task with caution, even though it may be simple and normally I'd do it easily. And when I have several successes, I get overconfident and commit to hard tasks (and sometimes even complete them, by a combination of experience, luck, and other factors). Musk had some great successes and believes it is his inherent feature. (Or, at least, this is how he appears.)

The good thing is that we, as humanity, benefit from his successes, but have also deal with his unfulfillable promises, not to mention occasional fits.

[tarsinge]

Not only sensors: in Europe I don’t see how driving in small towns could be possible without communicating with other drivers and pedestrians, and reading custom signs. Maybe add a robotic hand to the list of required hardware.

[anonymousab]

> Some Teslas don't have the sensor hardware and the compute power to do road-safe FSD.

Come on now. Elon was the one being pigheadedly stubborn about camera vision over LIDAR. There's really no reason to think this was a case where the engineers were insisting on a viable approach that unfortunately proved infeasible over time.

They weren't going to be able to do it with the tech they put in the cars and damn near everyone knew it. We shouldn't think their engineers were uniquely blind to what everyone was saying when it's fairly clear that there was a top down push here.

[cameldrv]

Even with advances in AI, I don’t think the cameras are good enough. They are prone to getting blinded by the sun due to poor dynamic range, and some of them have no way to clean themselves, and aren’t redundant for certain directions.

No matter how good the AI is, the car is not going to be able to drive if the image is a big white blob of blown out sun or bird poop, and there is no redundant sensor.

[philistine]

A car without radar cannot be trusted to drive itself.

I can move my eyes, put on sunglasses, lower the sun visor. That rinkydink camera is stuck looking straight at the sun. It’s a no go.