As a developer myself i hear this so often. Yes, it's (probably) your job on the line as well, BUT everybody nowadays can be a developer.
Maybe, don't call it developer. Everybody nowadays can built some software that fulfills some need (or just for fun). It's actually something good, isn't it?
Would you mind, if you could suddenly, over night, be a carpenter, a watch maker or glassblower?
My view is if you joined two boards together: congrats, you're now a woodworker!
Gating things by the level or complexity of what is accomplished isn't super helpful. I saw a lot of that with things like people that wrote HTML weren't "real coders".
Glassblowing and watch making require specialized equipment I suppose so if you can get your hands on the gear and try it out, I feel like you can call yourself that. Maybe throw a "hobbyist" tag on it if you feel inclined.
I like this mindset and I’m going to steal this quote. Everyone is figuring things on their own skill journey with their hobbies and careers. Wherever you are, you can learn more and help people to come with you.
I care a lot more about my gainful employment compared to the everyman being able to create their own... something.[1] That’s very selfish of me, AI Industry, and I’m sorry.
> Would you mind, if you could suddenly, over night, be a carpenter, a watch maker or glassblower?
Yes. If that came with the little-bitty stipulation of being tied to the “compute” property of these AI behemoths.
Of course going beyond that it gets less problematic. Some carpenter-skilled exoskeleton is a lot less problematic than disembodied intelligences spamming our minds and each other.
[1] What exactly? My mind is so saturated with AI tools being literally devtools. On top of devtools on top of devtools.
Well we can finally get back to a semblance of the status quo ante Internetum with something like businesses being able to list all of their contact information on their own websites instead of “see our FB page”.
Oh I know. But people often point to the API usage cost as an indicator of the magnitude of subsidisation, or to say that the big labs are far less efficient than cheap competitors.
I'm saying that this is not necessarily the case. They do a lot of optimisation and don't have the same price pressure to lower margins. They may not be losing as much on subscriptions as people think.
Oh hm, I've never seen this. API prices have always been exorbitant, I'm sure they're making good margins on that. Let's hope they aren't losing as much on subscriptions, because I'm not ready for everything to be API costs.
> Forget maintaining stability, AI is a downright destabilizer. It worsens understandability, fixability, debuggability, teachability, guaranteability, all the bloody bilities.
This is just an assumption and the whole article falls flat if this turns out to be wrong.
In my limited (as everyone else's) experience, working with agentic AI needs good documentation, good specification (spec driven, you know it's all the hype nowadays). Those alone lead to much improvement. Now take into account that probably your senior dev also has more time to think about the big picture, to improve all those little things that were a nuisance in the past but now are a mere "Claude, fix that" in a worktree away.. I would not bet on the assumption of this article.
I'd rather like to know if any real world usage broke, before coming to the conclusion that an edge case synthetic benchmark is worth changing the kernel (back or wherever) where supposedly the change that broke the benchmark had real world benefits.
Since we will never know it might be a good idea to feature gate the change, change the default and let users decide to change it back.
This may give some feedback on the lkml or else to decide if the change is worthwhile?
I am not questioning the benchmark.
But the benchmark is NOT measuring a real world application in a real world setting.
Anyway, I am merely wondering IF there is a company out there affected, at all.
I understand that this was only measured on a graviton 4 setting with very heavy lifting, without huge tables.
For example, this issue aside, I'd rather split such a workload into multiple smaller instances, naturally. Because the impact of a crash on this single node, heavy load, many cores, many clients scenario would be huge.
I recently had to switch to macos for work and Jesus Christ is this not the pinnacle of engineering. Sure, I'm accustomed to my self configured Linux desktop but boy is Mac OS slow to use and hard to configure.
And so keyboard unfriendly.
Would you mind, if you could suddenly, over night, be a carpenter, a watch maker or glassblower?
reply