The title confuses what’s actually happening. All of the senior devs left and now it’s a tiny, skeleton crew. Other than the growth of the code base (which is arguably feature creep and a sign of bad product design), this would have happened with or without AI. I do feel for him, but losing so many key people means nobody will understand their code regardless of.
I mean once you buy the tooling, it’s still probably cheaper to produce stainless steel trucks than paint normal steel. I agree on everything else though
If that’s true it would only be the case if you could amortize that tooling across enough vehicles. Selling 10k-20k per year from a production line planned with a capacity of 125k per year won’t get you there.
Not really. Stainless is expensive and brittle. It's a lot more work to form complex panels out of stainless, and then they're essentially scrap if there's even a tiny amount of damage. A small dent will start a crack.
The counter argument to that is well designed modern cars have great info-tainment systems which get regular updates. My Tesla gets regular software updates, has upgradable hardware, and won't be subject to these kinds of issues for decades. Old cars encouraged exchanging and needed aftermarket parts because they didn't care or try, so the products were low-quality and often failed.
> My Tesla gets regular software updates, has upgradable hardware, and won't be subject to these kinds of issues for decades.
Teslas from four years ago with Intel Atom processors and HW3 are already getting different and lesser updates than new Teslas, without any hardware upgrade plan in place.
A Tesla you buy right now is running HW4 which is inferior to current Cybercabs and HW5 has already been announced.
> well designed modern cars have great info-tainment systems which get regular updates. My Tesla gets regular software updates
And tracks you everwhere you go, resulting in that data being available to hackers, marketers, and any sleazeball that wants to extort you by suing you.
Not unique to tesla, of course, but a property of all those 'well designed modern cars'.
Legacy automakers are still using the "ship it and forget it" approach. If I want my 2023 Hyundai to have the same features as a 2025 (manual battery preconditioning, Google Maps directions on HUD, wireless Android Auto, etc), I need to root the head unit and install a bunch of vibe coded packages from randos on GitHub.
Although to be fair the vibecoded packages are probably higher quality than what Hyundai would've shipped.
Correct. Without root, the only way to heat the battery was if you use the cars built in nav to get directions to a fast charger
The problems being:
1. It didn't always recognize a destination as a fast charger
2. The built in nav is missing recently opened fast chargers
3. It only does this if you have the "preconditioning" setting enabled, but the car has a bug where if you use the (maybe only US version) Hyundai app to change the AC or DC charge limits it unchecks that setting.
Ah, I have an Equinox EV and it will automatically precondition based on the on-board maps, which, in my case, is Google Maps. I don't have the luxury of Android Auto or Carplay unless I jump through some hoops and self-publish an app in the Play Store. And even then it is janky.
Even so, I have a button in the battery settings which can start the preconditioning whenever I want to. It's been a fantastic car, but I am envious of your DCFC capabilities. But not of your ICCU. :-)
Is he not right though? Have you not interacted with a once star senior engineer who now operates at the level of a low-quality junior because they refuse to use AI?
There was a point a few decades ago with architecture where people moved off blueprints to CAD files, and everyone had to learn CAD or be pushed out, which lead to terrible churn and people I know got kicked out of architecture for good. This is a similar situation; they're not kicking out engineers for being 5% less productive, they're re-interviewing to find engineers who are 1/10th or 1/100th as productive as their peers. It sucks, and I feel bad for people who have to fear for their job, but it's a necessary action
I have yet to see data that supports this claims. In the contrary all I have read and observed point in the opposite direction.
Also I have yet to see an easy way to assess the performance of engineers. Lines of code it is certainly not!
This doesn't even make sense. You don't go from star senior to low quality junior by just doing what you've been doing and refusing to prompt the slop machine. If anything you're retaining skills that everyone else is losing. Prompting an llm is not a skill, the only thing this "re-interview" assesses is if you're using the magic box like the idiot CEO wants you to. Which is another way to say, are you de-skilling yourself fast enough, because that's what the CEO thinks is best for the company right now.
No, we are not there yet. Right now, AI is de-skilling the engineers who are using it.
Sure, they may churn out a lot more code, but what is the quality of that code? What is the security of that code?
Can you prove that the code they are generating is actually having a material impact on the bottom line?
If the AI tools suddenly became too expensive to use, could those engineers go back to doing things the way they once did?
If you want to vibe code, feel free to do that on your own personal projects. For everything else, you need to be able to demonstrate that what you’re doing is not just fluffing your lines-of-code metrics, or your tokenmaxxing metrics, and actually demonstrate real-world positive impact on your bottom line.
Maybe in a few years, we will get to a point where no one will need those skills anymore, and all that stuff will be done locally, so we don’t need any more massive datacenter projects that suck down all the power and all the cooling and create more problems than they solve.
But we are not there today. And the way that the AI bubble is already crumbling, I’m not convinced that we will get there any time soon.
> Is he not right though? Have you not interacted with a once star senior engineer who now operates at the level of a low-quality junior because they refuse to use AI?
I don't even know where to start. Let me just say it that by using LLMs you voluntarily decide to let the atrophy happen to your skills, knowledge and experience. You do things without understanding them, or with vague understanding only. Your thinking process becomes completely dependent on something that you find more and more difficult to verify (if you even care). Why would anyone do such a thing to oneself?
AI is maybe a 10% productivity increase for the entire job _if you maintain previous levels of quality_.
I see AI's main value-add right now as a ready-made excuse/bludgeon to finally let CEOs explore exactly how dogshit they can make everything while still having users.
<< Have you not interacted with a once star senior engineer who now operates at the level of a low-quality junior because they refuse to use AI?
I can honestly say no. I might question your premise a little. Lets hypothetically say you got Karpathy level senior; clearly a star and beam of sunshine in an otherwise dreary world. Is he now less of a star just because he won't use AI?
It is a real question, because if he is then the question becomes why.
Did he simply stop working as much? Or is he working the same amount, but it looks like he is doing less? If it merely looks like it, we may well have office situation on our hands, where rank and file may end up turning to management and asking themselves and the management one simple question:
What exactly it is you are doing here?
I know for a fact good management can enable some things, but I am starting to question the blind fervor I seem to see more of lately. If you can't motivate Karpathy or worse, you now are suddenly are unable to see his value, then maybe it is not Karpathy's fault.
<< This is a similar situation; they're not kicking out engineers for being 5% less productive, they're re-interviewing to find engineers who are 1/10th or 1/100th as productive as their peers
And this may be the big part: productivity. The problem is that productivity along the lines of silly metrics like: lines of code, agents commits per engineer are.. uhh.. not useless exactly, but, how do I phrase it a sensible way so that it reaches at least one stray manager that watches this conversation,
*really fucking dumb metric to treat as a measure stick.*
But who am I to judge.
<< but it's a necessary action
None of it is necessary. By definition it is all elective. That said, the nature of that relationship goes both ways and, I will be super subtle here, right now it is not the engineers, who can't do it all.
Quite the opposite, management could find itself in a pickle now, because it has neither skill to drive the development, practical knowledge that you only get from getting to close to the abyss, 'taste', theoretical foundation, curiosity or even basic exposure that would it let develop either of the previous facets.
Make no mistake, a lot of us choose to do deal with some of the idiocy for a stability gain. But, if there is none, why do I need you? I might as well try to strike out on my own and watch the AI coded towers fall ( being mildly dramatic for visual purposes; amusingly, I don't think I am that far off ) .
This was my thought. The last recession bailout was by democrats (in 2008), and they did so to influence the industry in their direction. I have a hard time imaging republicans would do a bailout, because they have no interest in creating new regulations even in their benefit.
In 2008 George Bush was president and he's the one who signed the bailout proposed by his treasury secretary. It would have happened under Democrats too no doubt, but saying it was done by Democrats and Republicans have no interest in them is just not true to reality.
While I agree in principle that most chips aren't made in the US, tons of 0805 capacitors and resistors are made in the US. Vishay does a lot of that here, as well as a few other manufacturers. Mouser lists country of origin and you can search that way.
This is great, especially as easypcbusa is being absorbed into a much more expensive conglomerate. Looking forward to seeing if they do just-pcb pricing as well
You say that but the economy seems pretty strong to me, we have lower than average unemployment with above average job listings. It's harder if you're on visa otherwise it's pretty competitive
reply