Considering he asked twitter programmers to print out their pull requests Im not even sure he’s not cosplaying a programmer

I completely agree that if there are tools that can allow a vehicle to “see” better than a human it’s absurd not to implement them. Even if musk could make a car exactly as good as a human, that’s a low bar. It isn’t good enough.

As for humans: if you are operating a vehicle such that you could not avoid killing an unexpected person on the road, you are not safely operating the vehicle. In this case, it’s known as “over driving your headlights”, you are driving at a speed that precludes you from reacting appropriately by the time you can perceive an issue.

Imagine if it wasn’t a deer but a chunk of concrete that would kill you if struck at speed. Perhaps a bolder on a mountain pass. A vehicle that has broken down.

Does Musk’s system operate safely? No. The fact that it was a deer is completely irrelevant.

Yeah. I mean, I understand the premise, I just think it’s flawed. Like, you and I as vehicle operators use two cameras when we drive (our two eyes). It’s hypothetically sufficient in terms of raw data input.

Where it falls apart is that we also have brains which have evolved in ways we don’t even understand to consume those inputs effectively.

But most importantly, why aim for parity at all? Why NOT give our cars the tools to “see” better than a human? I want that!

If you watch the video, the deer was standing on a strip of off coloured pavement, and also had about the same length as the dotted line. Not sure how much colour information comes through at night on those cameras.

The point here isn’t actually “should it have stopped for the deer” , it’s “if the system can’t even see the deer, how could it be expected to distinguish between a deer and a child?”

The calculus changes incredibly between a deer and a child.

To be fair, if they’re driven by an LLM I would still expect it to be wrong.

Sounds like they basically crafted some special messages such that it’s nonsense at 2.4ghz but smoothes out to a LoRa message on a much much lower frequency band (<ghz).

Even worse than it being wrong, is that by nature of the tool it looks right.

It reads to me like someone trying to build a persona that Elon would like.

If you’re deploying weapons on your own territory to reduce the operational capacity of an invading force then it’s by definition defending your country.

If you have a problem w/ this you’re going to have to cycle to the next argument because this one is nonsense. NEXT.

They will say, while simultaneously ignoring every NIST recommendation

If it depends on how ES6 is received they still have another 6 or 7 years

Somewhere in an ancient crypt, the bones of Luther begin to twitch to life…

I agree with everything you just said, but i think that without greater context it’s maybe still unclear to some why I still place chatGPT in a league of it’s own.

I guess I’m maybe some kind of relic from a bygone era, because tbh I just can’t relate to the “I copied and pasted this from stack overflow and it just worked” memes. Maybe I underestimate how many people in the industry are that fundamentally different from how we work.

Google is not for obtaining code snippets. It’s for finding docs, for troubleshooting error messages, etc.

If you have like… Design or patterning questions, bring that to the team. We’ll run through it together with the benefits of having the contextual knowledge of our problem domain, internal code references, and our deployment architecture. We’ll all come out of the conversation smarter, and we’re less likely to end up needing to make avoidable pivots later on.

The additional time required to validate a chatGPT generated piece of code could have instead been spent invested in the dev to just do it right and to properly fit within our context the first time, and the dev will be smarter for it and that investment in the dev will pay out every moment forward.

As a Sr. Dev, I’m always floored by stories of people trying to integrate chatGPT into their development workflow.

It’s not a truth machine. It has no conception of correctness. It’s designed to make responses that look correct.

Would you hire a dev with no comprehension of the task, who can not reliably communicate what their code does, can not be tasked with finding and fixing their own bugs, is incapable of having accountibility, can not be reliably coached, is often wrong and refuses to accept or admit it, can not comprehend PR feedback, and who requires significantly greater scrutiny of their work because it is by explicit design created to look correct?

ChatGPT is by pretty much every metric the exact opposite of what I want from a dev in an enterprise development setting.