abuani

I'll counter punching down on frontend engineers by positing that he was responsible for sandboxing their training infrastructure. Since clearly they don't know how to lock down an environment either without using llms

... Token cost is still heavily subsidized by the frontier labs while they're private companies and raise billions whenever they decide. The downward cost in tokens is unsustainable long term without revolutions training or inference of models.

> Here’s a third one. In some of the code reviews I’ve encountered that AI gives a lot of feedback, it’s just providing noise

Something I've found fun is seeing how long it takes for an llm review tool to come back satisfied with a PR. Think 100 lines of code changed, nothing terribly significant, but also not trivial. I'll have a local Claude session setup to babysit the PR and wait for feedback, accept all the recommendations, push the change up and request a review. I cap the number of iterations at 10 just so I'm not blowing a stupid amount of money. I've yet to come up with a PR where the llm reviewer is satisfied with the changes and has _no feedback_.

So where's the reasonable cutoff point for llm based reviews?

For me, private equity gets a bad rep because they've bought up all of the local mom and pop shops for just about any service I need thats face to face. All of the plumbers, electricians, general contractors, vets, grocery chains, coffee shops, and other shit were bought and had every ounce of revenue squeezed out of them. The result? All of the things I need locally cost 3-5x they did and the service is far worse.

And as a quick edit: none of these examples I have were of failing businesses. Each of the owners had their reasons, but none was due to it being necessary for survival. PE on hackernews tends to focus on software/bigger purchases, but so much of the damage in my opinion is on the local and smaller scale.

What about a $10k to $20k bill? Or better, 100%-200% of the employees salary? At a certain point in time, finance is going to care greatly about the ability to predict costs. Pre AI, predicting the cost an engineer was quite easy. Salary + benefits + licenses for software. CI systems that were unbounded were rolled up into opex and treated as a separate line item. Where does token spend land in this now?

Finance teams above all else value predictability. Token spend is the opposite of predictable. Eventually these two forces are going to collide and some part of the system will break. My guess is it's token spend, not finance allowing unbound spend on a previously predictable part of the balance sheet.

... Being gay does not limit altman or theil from getting married...

And last I checked, plenty of tech billionaires are married and by no stretch of the imagination stupid.

Reading this should make everyone in the US furious about the state of our health system

Maybe GitHub being popped for their own insecure by design platform, will cause them to reconsider growth at all costs. I know it's wishful thinking, but the amount of security incidents the past few years because of how actions was designed is wild. It would be great for them to finally recognize this and take ownership.

I'm looking forward to marketing folks doing oncall and support work for the features they're shipping.

I take a peak every month or so at spend for my company and notice more and more are consumed $1k in tokens a month and it is bewildering to me how. I use llms daily, and see anywhere from $200-$400 tops. This is using the most expensive models, in deep thinking mode. So I'm not a Luddite against the usage of them. I just can't figure how _how_ to burn that much money a month responsibly.

I genuinely challenge someone spending $5-$10k a month to demonstrate how that turns into $50-$100k in value. At a corporate level, I'd much rather hire a junior engineer who spends $100-$200/month and becomes productive then try and rationalize $100k/year in token spend.

Ah appreciate it. A year ago it was very clear when something was written by an LLM, but now you've gotta look for certain characteristics. I try not to infer to much, especially because llms are really helpful for non native English speakers to write faster.

I'd like to make it a bit more normalized to have public writing be transparent about if llms were used and how. That makes it quite a bit easier for readers to focus on the content instead of debating how something was written lol

Out of curiosity, can you point to specific sections that reel of AI? I read the article and didn't see anything that immediately stuck out, but maybe I need to start looking for different signals.

Costs were reduced to $0. Can't get better then that for a product that OpenAI had no clue how to monetize

Curious to see how they can turn moltbook into a money maker. Do they sell ads to agents?

I'm surprised the article or the source for it didn't dive deeper into the impact the changes to H1B have had on the tech job numbers. Mind you I'm not trying to argue for or against them, but I find it hard to believe the changes aren't contributing at least a small amount to the drop.

The other thing not mentioned is the impact on the end for ZIRP. Every tech company with a pulse over hired such a staggering amount during the pandemic. It's not surprising these companies are returning to reality and not hiring back to the same levels.

Others would say this is exactly what they voted for. Unfortunately it's all about perspective, and after a decade of passively consuming hn, it's obvious where the sites interest lies in terms of moderating content.

Heh, "vi=hx" was exactly how I forced myself to spend a week in helix. Just go for it

It's the first editor since probably sublime text that I've genuinely enjoyed. Useful without any configuration, and very easy to get a productive environment.

There's a few rough edges that I'm trying to work through. I've been able to solve my "open in X" like key bindings. But I have yet to get things like "run test for current method". That's probably the biggest pain point I've had so far

Just to reiterate the point the person above you made, but in far simpler terms: independence can be far greater return on your personal well-being then maximizing gains. I'm willing to "lose" out on $50-$100k over the lifetime of my mortgage in exchange for never needing to make a payment on the house again

There will be a court case where some bit of evidence is going to be similar to:

Ice Agents: "Is <name> here illegally?" AI prompt: "you're absolutely right!"

Yeah I agree, we should be holding the ruling class accountable for building generational wealth while average Americans can barely afford basics.

And that attitude is exactly why humanities is a subject every computer science student should take.

I had a recruiter reach out for a company doing Agentic SRE's due to my years of experience as an SRE. Second sentence was describing their mission as making the SRE role no longer necessary for companies. I know if you read between the lines that's the goal of many AI companies, but I was surprised how upfront they were.

... That makes no sense. The value of an instrument isn't what makes a successful musician successful. The best musicians can make the cheapest sounding instrument sound amazing. What cheaper instruments offers is the opportunity for someone to even have a chance.

There are still a select few subreddits where this is true as well. I genuinely miss 10 years ago getting into random shit like double edge razors, home brewing and woodworking and how supportive those communities were to get into. Some communities _do_ exist, but once they get past a certain size it becomes worthless

Can it though? Everything I've seen and experienced is that LLMs are very good at making it appear to do those things, but the amount of times I've gotten stuck on "you're absolutely right!" When correcting the LLMs suggests that it can not reason by any means, nor does it learn. Otherwise, an LLM would never get stuck in a loop.

This was a very challenging article to read. Not because any of the concepts described, but for the way ideas are thrown around and organized. This looks like it was written by a set of llm agents that were instructed to write an article without a clear outlined, and then the author took what they felt were the best bits and hit publish.

Did I suggest that? I'm pointing out the blaring hypocrisy of a company sitting on $350M in cash that opted to double the size of their company without having a clear strategy to become profitable. Then after laying off half the company, the CEO publicly states it's because the laid off workers don't have the skills they need for "the new era". I would really like to see in these scenarios the CEO accept a tiny bit of responsibility for their failures to set strategy and over hire, instead of publicly shaming 70 people they chose to hire in the first place. That's a failure in leadership, not in employees.

That sounds a lot like the CEO is throwing 50% of the company under the bus for his own failures to you know, make the business viable.

A counter example would be Weaveworks(folks behind Flux/FluxCD and many other widely used oss tools). I'm sure the ex employees would've preferred to get acquihired vs closing up for good. I highly doubt Styra was pulling in enough money to fund their business, and the days of zirp are long gone, so I doubt they would've been able to raise another round to keep the lights on for another few years.