It seems right that LLMs don't have an innate understanding of time, although you could analogize what you did with writing someone a letter and saying "please count to ten with a two-second pause between numbers". When you get a letter back in the mail, it presumably won't contain any visible pauses either.
That's because you used a LLM trained to produce text, but you asked it to produce actions, not just text. An agentic model would be able to do it, precisely by running that Python code. Someone could argue that a 3 year old does exactly that (produces a plan, then executes it). But these models have deeper issues of lack of comprehension and logical consistency, which prevents us (thankfully) from being able to completely remove the necessity of a man-in-the-middle who keeps an eye on things.
just because it doesn't do what you tell it to doesn't mean it's not intelligent. i would say doing something that gets you where you want when it knows? it can't do exactly what you asked for (because architecurally it's impossible) could be a sign of pretty intelligent sideways thinking!!? dare i say it displays a level of self awareness that i would not have expected.
Industrial espionage isn't magic. Airbus once stole basically everything Boeing had, but that doesn't mean Airbus could magically build a better 737 tomorrow.
China steals a lot of documentation from the US but in a tech forum you of all people should be very familiar with how little actual progress a bunch of documentation is towards a finished unit.
The Comac C19 still uses American engines despite all the industrial espionage in the world because most actual engineering is still a brute force affair into finding how things fail and fixing that. That's one of the main advantages SpaceX has proven out with their "eh fuck it, just launch and we will see what breaks" methodology.
Even fraud filled Chinese research makes genuine advancements.
Believing that China, a wealthy nation of over a billion people, with immense unity, nationality, and a regime able to explicitly write blank checks could only possibly beat the US at something by cheating is like, infinite hubris. It's hilarious actually.
I don't know if DeepSeek is actually just a clone of something or a shenanigan, that's possible and China certainly has done those kinds of things before, but to think it's the MOST LIKELY outcome, or to over rely on it in any way is a death sentence. OpenAI claims to have evidence, why do they not show it?
>>>Believing that China, a wealthy nation of over a billion people, with immense unity, nationality, and a regime able to explicitly write blank checks could only possibly beat the US at something by cheating is like, infinite hubris. It's hilarious actually
So this is the first time I’ve heard the Chinese regime being described in such flowery terms on HN - lol. But ok - haha
My assumption was always that PG's site was just hand written HTML (really just the same template with new p tags) with a link added to the essays page.
1. The most popular front-end frameworks (e.g. React and Angular) are created by large orgs which naturally leads to extensive codebases as they’ll need to have many use cases.
2. Every new and shiny front-end framework is good for job security since having it on a resume makes you more hireable. I suspect it’s also why some developers practically force startups to switch their front-end codebase to the next shiny framework - resume padding.
As a solo dev I wasted a few years using (and hating) these new frameworks for my Django projects because it’s what everyone was recommending.
Then it hit me when I realized that they were meant for large dedicated front-end teams to increase their productivity. If you’re a solo developer, they actually hurt your productivity.
That’s why I’m so happy with htmx. I can get my sanity back!
Well, many open source developers are quite content with making their software freely accessible without expectation of monetary gain. In fact some are strongly opposed to it (see uBlock Origin maintainer). And that’s ok too!
Honest question: how valuable is the current mode of testing (based rote memorization) if much of that can be done by AI anyway? Maybe it’s time for new testing methodologies?
Where are people testing just rote memorization? At least 20 years ago when I was in lower school, most of what we did had only a very small memorization component. The SAT, ACT, SAT subject exams, etc only barely rely on memorization, and many questions have no memorization component at all. It’s why 4th graders can get top-percentile scores on some of these.
You need both memorization and applied reasoning to be taught in school. Without something to apply reasoning to you can’t really reason about anything. And without applied reasoning, memorization is pretty useless.
I think it's fine for ai to be able to answer test questions?
If the ai can get a good answer, the questions must be fairly well constructed, and it's reasonable to expect a student exposed to the same info could also come up with a good answer.
All great points. To add:
As someone who's worked blue collar jobs, it is all too common for unions to protect assholes from getting fired. This leads to a hostile work environment created by the "high seniority" workers.
Something I don't like whenever these discussions come up is the condescending tone, from white collar workers. "Don't these poor people know what's good for them????"
Working class people are capable of thinking for themselves and it's not that uncommon for people to move from a union shop to a non-union shop due to the reasons outlined above.
All the things you dislike about unionized working environments are things that can go wrong in any working environment. If we had more unions, you'd have more of a choice, instead of the situation now where only the most corrupt survive.
Poor people? I think rich techies are crazy - or rather, badly misinformed- not to start a union.
Exhibit A of poorly informed: grandparent poster seriously thinking he has more individual bargaining power against trillion dollar corporations than a collective bargaining agreement would.
Exhibit B: grandparent poster thinking collective bargaining _must_ entail many aspects like seniority based compensation that are totally optional.
Tom Brady is in the NFL players association. Tom Cruise is in the Screen Actors guild. Naive techies think being good at leetcode hard gives them bargaining power. A lot of ex Tweeps are getting a wake up call.
The only reason techies had the illusion of negotiation leverage is that they were generating such absurdly high revenue for their employer, the slice they got seemed huge compared to the rest of the country’s gutted middle class. Factor in inflation, housing, and that Wall St is now gunning hard to bring tech compensation down, and you’ll realize techies are the last gasp of Americas dying middle class, and that really pisses off activist hedge funds . See Elliott Managements recent takeover of Pinterest so they could “re-level” employees.
Read the emails between Steve Jobs and Sergey Brin w.r.t high tech class action lawsuit. Read about Google hiring union busting consultant firms.
If unions didn’t work, tech execs wouldn’t be so terrified of them.
Yes, poor people should also unionize. And yes, unfortunately a lot of their options are as broken and corrupt as their employers.
That doesn’t change the fundamental fact that collective bargaining is the _only_ answer .
> Read the emails between Steve Jobs and Sergey Brin w.r.t high tech class action lawsuit. Read about Google hiring union busting consultant firms.
Worse still is that these firms have been caught collectively bargaining against workers through illegal secretive non-compete clauses. Wage suppression has been going on for years, even while these companies have been at the very top of thriving businesses.
It’s simply not a question of affordability. These companies did it because they could.
I’d also like to point out that much of the union activity of a hundred years ago was also about poor working conditions, and not just compensation. Much of that was quelled by increasing government regulations that protected workers against exploitation. Is the government adequately performing this job?
> Tom Brady is in the NFL players association. Tom Cruise is in the Screen Actors guild.
You think that has anything to do with their compensation? Tom Brady has won how many Super Bowls? Tom Cruise has had how many blockbuster movies?
The opposite should be true if unions were the great equalizers, no single actor/athlete would be making hundreds of millions a year and all of them would be making enough to buy a house in Beverly Hills.
> The opposite should be true if unions were the great equalizers, no single actor/athlete would be making hundreds of millions a year and all of them would be making enough to buy a house in Beverly Hills.
This makes no sense. Two people that are in unions are doing the thing you say they can't. What in the world?
When I tell an LLM to count to 10 with a 2 second pause between each count all it does is generate Python code with a sleep function. Why is that?
A 3 year old can understand that question and follow those instructions. An LLM doesn’t have an innate understanding of time it seems.
Can we really call it AGI if that’s the case?
That’s just one example.