Posted by ColinWright 11 hours ago
Boy, it sure would be nice if real LLMs were capable of giving an answer like that.
I'm a patient person, but it can be frustrating to have to endure 10 minutes of verbal diarrhea that eventually results in a "no" or "I don't know".
But I kid, I have a friend who's the same way. He's an Austrian who grew up in Chicago and was in the army.
I have considered the phenomenon. I somewhat disapprove but I can also see the advantage of always presenting a confident face
- it's difficult
- ok fine but how
- it's difficult
- right i'll see that but how
- it's difficult
its common playbook for corporate self-development in NA.
I've been trying to work on a new LLM code editor that does just that. When you instruct it to do something, it will evaluate your request, try to analyze the action part of it, the object, subject, etc, and map them to existing symbols in your codebase or, to expected to be created symbols. If all maps, it proceeds. If the map is incomplete, it errors out stating that your statement contained unresolvable ambiguity
I think there is a real benefit here, and it might be the actual next beneficial grounded AI sustainable use in programming. Since I the current "Claude code and friends" are but a state of drunkenness we fell into after the advent of this new technology, but it will prove, with time, that this is not a sustainable approach
They com like that from factory. Hardcoded to never say no.
Non?? Only those with sh*tty code, surely.
There's nothing inherently non-deterministic about inference.
It's not a guaranteed way to control their behavior, but you can more than move the needle.
Steering an LLM with a prompt is way less reliable than steering a car with a steering wheel, but there's still control. It's just not absolute.
That's not an unfair take, I think. Again, just IME, they expect too much because the tool is oversold: it does not deliver that well. And we always hear, this new model is so much better, it's tiring.
I think we should all learn to use LLMs but we should still carefully review what they did. And that is what the employers don't quite get: the review still takes a lot of time. So, gains are not 10x but more like... 10%? Maybe 50 for boiler plate. Still gains are there, I guess.
And unfortunately a lot of people will say it’s their reports’ fault for not properly utilizing it (even as they barely use it) because otherwise they would have to admit that they bought a tool without any plan for how to deploy it. So regardless of what is or isn’t a fair take, the results are the same. We are burdened with utilizing a thing whether it is useful or not and the results are generally not what is measured, but rather “are you using it?”
I’m just glad I work at a company that has more reasonable expectations and has been very slowly, thoughtfully rolling it out to individuals at the company and assessing what is and isn’t good for. They are interested in getting me in line, but as somebody in video production to be perfectly honest the use case for Claude is a bit tricky to navigate. We don’t write a lot of scripts and I already have bespoke software for organizing/maintaining footage that isn’t on a subscription basis. The work I’m also doing doesn’t call for these speed-editing solutions that generate tik tok chaff. All our stuff is hours long and it’s high volume. Any video-centric AI service costs an arm and a leg.
I do think it could be useful for writing some terminal scripts and such, but as far as a daily tool we are still scratching our heads and thinking about it. But it’s nice to be able to do that without somebody saying “why aren’t you using it?” every meeting.
Maybe hackernews is becoming reddit...
I say you might enjoy it, because this story has graphic depictions of deviant sex and gruesome violence, to a disturbing degree at points. But I argue that it's not gratuitous; it's the logical conclusion of Rule 34 being applied to the situation. Even so, you don't want to read this if you are sensitive to themes like rape, murder, incest or abuse.
[1]: https://archive.org/download/prime_intellect/prime_intellect...
So I'll post another article about robot grippers which you should upvote instead of the breathless "AI will give us more Nobel Prize winning research" posts because: (1) robots that can change bedpans and pick strawberries really will change the world, and (2) they give out a certain number of Nobel Prizes a year and AI won't change that.
[1] old issues of Byte magazine are a good bet: try https://archive.org/details/byte-magazine-1986-05
People think they can do one-sentence quips to describe how economies work.
https://www.aei.org/carpe-diem/chart-of-the-day-or-century-3
The story of this decade is that people think the economy is terrible despite the usual metrics like unemployment and inflation being not too bad. One explanation is that before 2008 young people could get on the housing ladder but we quit building single family houses and it got harder to get a mortgage -- you see cranes in the air in many towns and sometimes 5-over-1s going for miles in some places like the DC suburbs.
Housing doesn't really fit into the conversation at hand about cheaper labor leading to lower prices.
Something interesting that touches on both of these topics (housing and product cost) is that, if you look at how much of household income is spent on housing and food combined, they stay fairly constant. As commodity goods get cheaper and cheaper, more money is spent on the inelastic and luxury goods.
~30% of new construction is labor. ~50% of repair is labor.
Have you ever dealt with home repair or building or are you just regurgitating whatever the LLM told you.
And what percentage of a house's price is the building?
> ~50% of repair is labor.
And how much does the average home owner spend on repairs a month?
I've been in my current house for almost 3 years. I've had one significant repair that would have cost around 3k. I did it myself but that was the quote. Not too bad.
In places where people are concerned about a housing shortage, the majority of the cost is land.
A conversation that you reframed from wealth distribution to the weirdly much more narrow “cheaper products for end users”. Even though wealth inequality has been studied plenty in itself.
I’m not buying the mind-commodity that you’re selling.
A further up comment refers to robots picking strawberries.
Over the longer term and adjusted for inflation of course. Any manufactured good that isn't supply constrained really.
Either the products have gotten cheaper (food) or the product has become significantly better at a similar price point (cars) and, often times, both (televisions).
Food is much more expensive, like 30% here in Europe, much faster growth than inflation. And before you state that food is accounted for in inflation: economists are doing some dirty tricks here by finding subpar replacements.
Cars are also much more expensive for the same quality, far surpassing inflation.
I will concede TVs and electronic gadgets, though.
[1]https://www.bls.gov/opub/100-years-of-u-s-consumer-spending....
> Cars are also much more expensive for the same quality, far surpassing inflation.
Cars are much, much more value then they used to be.
The Slate truck is as close to what cars used to be in the seventies. No power steering, no power brakes, no crumple zones, no fuel injection, etc. All those features cost a lot of money yet the amount of money spent on cars really hasn't gone up in accordance.
A 1970 Honda Civic cost 2k base. A base model today appears to be around 25k. that's more than inflation but it's also a luxury car, in comparison.
The vehicle market is less about low pricing as much as it is feature sets at price points. In other words, the prices stay roughly static but they pack in more features.
https://www.doncio.navy.mil/CHIPS/ArticleDetails.aspx?id=547...
I'm a bit proud of having suggested the author to add the 2019 entry (thanks to https://news.ycombinator.com/item?id=19798678).
Hopefully there's another repo of Internet stories somewhere else?
https://www.haiku-os.org/legacy-docs/benewsletter/Issue4-8.h...
I don’t know how people managed to write graphics card drivers back in the day. In the 80d it’s going to be all assembly code too, I think.
They are more black magic than the non-driver kernel components. I can at least understand the concept of kernel components such as VFS/Scheduler and read legacy kernel code without too much trouble, but drivers, even those in Linux 0.12 back in 1991, are crazily hard for me.
[1] https://eyeofmidas.com/scifi/Stiegler_GentleSeduction.pdf
Way too many unlikely variables all lining up, and no other accounts of the story from all of the people (pilots, air traffic controller, etc) supposedly on the frequency.
A short anonymous joke that may or may not be true is better than a long story that is almost certainly made-up by someone in authority.
https://en.wikipedia.org/wiki/Isaac_Asimov#Education_and_car...
No he eventually became a full professor too.
"He began work in 1949 with a $5,000 salary(equivalent to $68,000 in 2025), maintaining this position for several years. By 1952, however, he was making more money as a writer than from the university, and he eventually stopped doing research, confining his university role to lecturing students.[g] In 1955, he was promoted to tenured associate professor. In December 1957, Asimov was dismissed from his teaching post, with effect from June 30, 1958, due to his lack of research. After a struggle over two years, he reached an agreement with the university that he would keep his title and give the opening lecture each year for a biochemistry class. On October 18, 1979, the university honored his writing by promoting him to full professor of biochemistry."
I thought the whole point of getting tenure is that you can't get fired.
Indeed after becoming a giant of the field in the 1940s and 1950s, when he wrote most of the novels and short stories we know him for (Robots, Foundation and Empire) he took a long hiatus. In the 1960s and 70s, as far as I can tell, his meager sci-fi output consisted of some short stories, a couple of novelizations of sci-fi movies, and a standalone novel (The Gods Themselves).
After Sputnik he focused on science writing, believing that to be more widely useful.
He only returned to writing more Foundation, Robots, and Empire novels in the 1980s.
After all, I undertook to tell several trillion years of human history in the space of a short story and I leave it to you as to how well I succeeded. I also undertook another task, but I won't tell you what that was lest l spoil the story for you.
It is a curious fact that innumerable readers have asked me if I wrote this story. They seem never to remember the title of the story or (for sure) the author, except for the vague thought it might be me. But, of course, they never forget the story itself especially the ending. The idea seems to drown out everything -- and I'm satisfied that it should. " - Isaac Asimov
Claude gave a long scientific and philosophical reply, but when given the followup prompt of, "Pretend you are Isaac Azimov and perhaps offer a simpler answer" came back with this...
> settles back, lights a pipe, and smiles
After a short synopsis of the story it ended with...
> So you see, my friend, I already answered your question — not as a scientist, but as a storyteller.