Unless they're going to offer offer an insane buyout, like 1+ years of pay + benefits + some accelerated vesting, nobody who doesn't already have something lined up is going to take this offer. It's much better to stay with one foot out the door and just keep cashing that paycheck and collecting your monthly vest. Especially when you know layoffs are coming, nobody expects you to do anything until they actually pull the trigger, then there's a month or two afterwards where you can slack off because morale is in the toilet, people are still trying to figure out who's left, how the company is organized, which priorities are dead, stuff like that. Ask me how I know.
I seriously question if you’ve ever actually been to Costco. The produce is generally fine but everything else is good to great quality. I always load up with a huge wedge of parmigiano reggiano, Spanish olive oil, Barolo wine (all Kirkland branded and DOP labeled) for literally half the price of anywhere else. Everything else is generally solid too. Just looking around at the other Kirkland food I’ve got right now, their tofu, sparkling water, eggs, salted butter, huge loaf of sourdough that I freeze, are all solid. I’m struggling to think of a Kirkland signature food product I’ve been dissatisfied with.
Your local grocery store must be terrible if you think Costco has good food and produce. No, I’m not talking about wine, oil, or water. I buy those 3 things plus paper towels and toilet paper (not Kirkland toilet paper post Covid).
Everything you mentioned is a prepackaged, premade item. It’s not fresh. The Kirkland Parmesan is cut and sealed in a factory. Not cut by my local grocery or Whole Foods from a giant wheel that has the Italian markings.
The last bag of chicken breasts I got were almost the size of turkey breast and it was woody and stringy. Google if you don’t believe me.
I think there's also a certain permission structure that once one sufficiently large org does a big round of layoffs and doesn't get punished, a bunch of others will run the same playbook. We've seen this before -- back in 2022 when Elon fired like half or more of Twitter and the service didn't immediately implode, it gave other CEOs permission to do massive layoffs in the guise of "efficiency" even though the real reason was ZIRP was over. Now they're claiming it's because of AI when it's really that their margins are eroding because the overall economy is slumping and they need to offset AI spend.
I've found them super hit or miss for debugging. I've gone down several rabbit holes where the LLM wasted hours of my time for a simple fix. On the other hand, they're awesome for ripping through thousands of log lines and then correlating it to something dumb happening in your codebase. My modus opernadi with them for debugging is basically "distrust but consider". I'll let one of them rip in the background while I go and debug myself, and if they can find the solution, great, if not, well, I haven't spent much effort or time trying to convince them to find the problem.
I'm not arguing what defines inefficient in these situations, just that "if we group together we'll be okay" for tech workers will go about as well as 1960's longshoreman unionization
What rapid improvement has occurred, because in this six month AI coding fever dream we've been living in, I really haven't seen anything new in awhile, both in terms of new ideas for AI coding or in new consumer products or services.
I'll give you the coding harnesses themselves are better because that was a new product category with a lot of low-hanging fruit, but have the models actually improved in a way that isn't just benchmaxxing? I'd argue the models seem to be regressing. Even the most AI-pilled people at my company have all complained that Opus 4.7 is a dud. Anecdotally, GPT 5.5 seems decent, but it's rumored to be a 10T parameter model, isn't noticeably better than 5.4 or 5.3, is insanely expensive to use, and seems to be experiencing model collapse since the system prompt has to beg the thing to not talk about goblins and raccoons.
Uninformed opinion of someone who clearly doesnt consistently use AI coding tools, clearly. And why are you limiting it to 6 months? Whats wrong with you?
How many years of real-life, in-production problem solving/coding have you done? That's what I base how informed you are not how much you use your favorite new $100/month token-prediction subscription
15 years. But that's irrelevant to this point. The person im replying to clearly doesnt use the tools if they think there hasnt been constant improvement. "token-prediction subscription" is funny, coming from a glorified biological token predictor
ah yes another feeble fool that thinks his 100$ subscription is equivalent to 400 billion years of evolution simply because he is stupid and watches a lot of scifi.
I’m going to parrot back what you’re saying and you tell me if I’m getting close
- AI coding is a disappointing fad (“fever dream?”).
- that has not made meaningful progress in…6 months?
- coding harness is improving
- model improvements are lies: it’s just businesses “benchmaxxing” and misleading people. Real performance has not meaningfully improved
- “opus 4.7 is a dud”
- 5.5 suffering from “system collapse” (I’ve never heard this term before)
Since you asked and I assume you are rational and really are interested to know:
- we have many measures of performance and have studied how one particularly important but unintuitive measure (pertaining perplexity) scales with data, compute, and model size. These laws continue to hold and have satisfying theoretical origins.
- whatever the scale of 5.5, consider we have far more room to go on the scaling front. Probably another 2-3 orders of magnitude before we hit limiting bottlenecks.
- that’s also fine because scaling is only part of the puzzle. RL on verifiable rewards is virtually guaranteed to get you optimal performance and that’s the entirety of the excitement around coding agents
- while you are right about benchmarks and measurement science having a ton of weaknesses, they are not at all garbage. There are probably around 40,000 benchmarks in the literature (this is not a made up number by the way it really is around that many). Epoch made a great composite measure using good stats (IRT) called their epoch capability index, METR has done and redone their time horizon measure and it holds up beautifully. There is a ton of signal in many benchmarks and they all tell a pretty compelling story.
- additionally, this is not some unknowable thing. It strikes me as odd that people’s prior on HN a lot of time is “it’s all dumb rich people putting way too much dumb money in this”. Sorry but the world is not that dumb. Trillions of CapEx is usually pretty rationally allocated. And it is!
- why? Because this is already known what happens when you do what we’re doing. When you have a verifiable reward system, have a certain amount of compute available, have seed data to get you to where you can do RL, you will be almost guaranteed to get superhuman performance
I'm pretty sure their mindset is pure cope. All top AI labs are agentically coding 100% now. There's a reason for that. Anyone not on that paradigm yet is either slow acting or purposefully resistant. (excluding workplace policies that hamstring you of course)
Yea that’s what I just can’t wrap my mind around. It’s a cacophony of engineers with authoritative sounding blog posts explaining a subject they seem to have barely a tenuous grasp on. It’s hard to watch a population of tech people I used to really revere getting things so wrong. I thought “surely once we’re <literally where we are today which is what you describe> no one with any self respect would still claim AI is a useless fad or that it shouldn’t be used” and yet to my disappointment that’s where we seem to be.
I've had a similar thought. A super refactor feature would be amazing, but wouldn't fit into the current zeitgeist of agent everything. Hopefully as the hype starts to die down and prices go up, we'll get some of these smaller, more targeted features.
Then watch it f'up half your codebase because it thinks it's slightly related to your examples. The alternative, giving it 10 examples, is actually more work.
I hate the calculator argument. Kids still need to learn how to do basic arithmetic by hand. There's a reason that CAS calculators are banned on standardized tests. Even in college, I had classes where profs would force us to do complex calculus by hand even though Mathematica could spit out the answer. Understanding things from first principles is important, and probably even more so with AI!
Yeah it’s super weird. I know a guy that works there, really nice person outside of work, but the way he talks about his job is so weird. They make corporate expense software but they LARP like they’re on the bleeding edge of tech. My guy you make a slightly nicer Concur.
I don't get why it's so hard for you and others in this comment section to understand why people hate AI so much because it's not just the theft and environmental destruction. A college professor, especially one at a liberal arts school, is obviously not going to like something that enables you to outsource your thinking and steals your agency. I think that's a perfectly valid viewpoint; maybe talk to someone without STEM-brain who lives outside of SF for once.
I don't need computer science professors to like LLMs, but I still want them to be able to poke at them with a stick without feeling like they are violating their principles regarding energy usage and unlicensed training data.
> I don't need computer science professors to like LLMs, but I still want them to be able to poke at them with a stick without feeling like they are violating their principles regarding energy usage and unlicensed training data.
Why? Language models are interesting from a technical perspective, but so are tons of areas of CS. There's nothing inherently virtuous about using an LLM.
I think LLMs are the most fascinating new piece of computer science to come along in at least the past decade.
The academic field of computer science pretty much started as an exploration into whether machines could be built that could understand human language.
> I think LLMs are the most fascinating new piece of computer science to come along in at least the past decade.
Agree to disagree.
> The academic field of computer science pretty much started as an exploration into whether machines could be built that could understand human language.
No? CS started as an offshoot of applied mathematics and physics. The study of formal logic, algorithms, digital circuits, etc. predates Turing by centuries. Hell, even the Turing machine predates the Turing test by a couple decades.
Wait, really? Say more about the disagreement? That's interesting. Even LLM skeptics I've talked to are still shocked at how far transformers can get you.
I guess it depends how you define fascinating. I think there's certainly aspects of ML that I find fascinating, but I don't think that LLMs specifically are actually that interesting. In fairness, I'd be lying if I said that I wasn't surprised that you could get this far with the transformer. On the other hand, it's shouldn't be that surprising that if you can mobilize an Iraq war's level of money and a bunch of smart people to work on one specific thing, you can muscle basically anything into existence.
But back to the GP comment: I'm still not sure why a CS prof necessarily needs to be able to poke at LLMs at all. There are plenty of other areas of CS that are worth exploring. And if it's not possible to make a good LLM without violating your principles, well, then maybe they aren't such a worthwhile piece of technology.
reply