Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Let's be honest, Generative AI isn't going all that well

garymarcus.substack.com

Post reply on HN
Print view View original

341–346 of 346 posts

Re: Let's be honest, Generative AI isn't going all that well

#341

Post by ta9000 »

cadamsdotcom wrote: post #235
mattmaroon wrote: post #6

Meanwhile, my cofounder is rewriting code we spent millions of salary on in the past by himself in a few weeks. I myself am saving a small fortune on design and photography and getting better results while doing it. If this is not all that well I can’t wait until we get to mediocre!

G’day Matt from myself another person with a cofounder both getting insane value out of AI and astounded at the attitudes around HN. You sound like complete clones of us :-) We’ve been at it since July and have built what used to take 3-5 people that long. To the haters: I use TDD and review every line of code, I’m not an animal. There’s just 2 of us but some days it feels like we command an army.

As a software engineer this scares me from an employment perspective but I also use Claude Code to produce 90% of the code I commit now (after reviewing and revision of course) so there’s that… ;)

Re: Let's be honest, Generative AI isn't going all that well

#342

Post by Ygg2 »

_heimdall wrote: post #336
Ygg2 wrote: post #262

Earlier quoted context omitted.

Sure, but all code is a potential future debt. You wrote a music player that only allows one artist from list of all artists? Tech debt. You wrote optimized assembly for x86_64? It's the year 2060, and we only support NGPU_ARM_N_LEG. The moment your expectations change (which is all the time), your code needs to be changed, and effort isn't free.

Tech debt is not part of a financial account or disclosure though. Yes those are forms of debt, no they aren't financial debts or financial liabilities.

Sure it isn't. It is still a financial liability. You need to sacrifice time clearing it to get to work that actually pays the bills.

Re: Let's be honest, Generative AI isn't going all that well

#343

Post by starchild3001 »

How to read this brilliant blog post by Gary Marcus?

- Replace all the occurances of "LLM" by "human";

- Replace all the occurences of "Scaling" by "additional education".

Voila! You get an article that actually makes sense, plus you'll get a better sense of where the technology is -- these models are behaving much like humans across many tasks. They aren't perfect. But they are getting better everyday, and are quite useful.

Thank you AI developers and researchers for making progress everyday! No "thank you" to people like Gary Marcus who'd be called a "perma bear" in financial parlance.

Re: Let's be honest, Generative AI isn't going all that well

#344

Post by kunley »

mattmaroon wrote: post #328
kunley wrote: post #258

Earlier quoted context omitted.

Millions? Of people in self driving teslas? The actual number of such vehicles produced is two orders of magnitude less.

No it isn’t. They sold 1.6 million vehicles last year and every one of them can self drive. At least once a year they give out free access to it and I don’t know a Tesla owner who hasn’t tried it. You’re right that it doesn’t happen every day but that doesn’t change the point, you all are debating whether something can happen after it already happened.

What has already happened is that you unfortunately have manipulated the facts. How long were you planning to not reveal what you really mean: that some drivers have an oportunity to use self-driving once year and that they allegedly tried it, althought that's really just opinion based on your own preference. That's much different than claiming millions of such drivers roaming everywhete at any moment of time.

Plus, the idea that every single Tesla has this option is just your rought estimate, right?

Re: Let's be honest, Generative AI isn't going all that well

#345

Post by hattmall »

immibis wrote: post #333
jimbokun wrote: post #301

Earlier quoted context omitted.

China is not surpassing the US in all aspects of technology. There is still much for them to steal.

Can you name one area of technology where the USA has better technology than China? I'll wait.

UV Lithography, Space Exploration, self driving cars, weather prediction, industrial safety, circular knitting, medical imaging, robotics, bioengineering, agricultural yield management

Re: Let's be honest, Generative AI isn't going all that well

#346

Post by moonshotideas »

joquarky wrote: post #314
moonshotideas wrote: post #211

How long do you think it will be until the “ai isn’t doing anything” people are going away 1 month, 6 months, I’d say 1 Year at the most, anyone who has used Claude code since Dec 1st knows this in their bones, so I’d just let these people shout from the top of the hill until they run out of steam… Right around then, we can send a bunch of reconnaissance teams out to the abandoned Japanese islands to rescue them from…

I think the "AI isn't doing anything" crowd have some kind of vocabulary/language barriers/deficiencies that prevent them from refining their prompting methods into something that works for them. I find that the more precise I am in my prompts, the more precise the response. But that requires that I use vocabulary that I wouldn't use in a human conversation.

Same here, I created a prompt enhancer gpt and a prose enhancer gpt, and I tend to chain all my prompts through them, then I use an extension to remove markdown + replace Unicode, and then add tabs and proper formatting to product a final version of all my prompts. This tends to result in prompts that perform 20-25% better for all difficult or multi-part tasks.

Logically this makes sense, the probabilites for next tokens the model produces follows the pattern it observes from the initial input, if your prose reflects that which individuals with higher intelligence tend to use, the model will continue this high in response and vice versa

I’m curious to see what happens as the % of training input using synthetic data generated by models tips the scales so that markdown reflects higher intelligence inputs - I wonder when this will occur

Post reply on HN