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

Zuckerberg says AI agent development going slower than expected

reuters.com

Post reply on HN
Print view View original

281–290 of 661 posts

Re: Zuckerberg says AI agent development going slower than expected

#281

Post by liuchao-001 »

Actually, agents have been developing incredibly fast. It is just that Zuckerberg has some unrealistic expectations; the man is a lunatic.

Over the past six months or so, OpenAI's internal team has completely shifted from being heavy ChatGPT users to using Codex. Once you start using an agent like Codex, it is very hard to go back.This shift is truly transformative.

I am also aware that some of the consumer agent products on the market are growing very rapidly, such as Manus and GenSpark. Not to mention Claude Code and Codex.

Re: Zuckerberg says AI agent development going slower than expected

#283

Post by thisisit »

efficax wrote: post #115

I was worried this time last year that by this time this year, companies would have slashed their engineering teams down to a handful and everything would be driven by mostly autonomous agents with human guidance. But it just hasn't happened. Do I write all my code with an agent now? Yes. Can you just give an agent a desired outcome and let it work, unsupervised? Absolutely not. I can produce more code than I used to…

> I was worried this time last year that by this time this year, companies would have slashed their engineering teams down to a handful and everything would be driven by mostly autonomous agents with human guidance. But it just hasn't happened.

Amara’s law: We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.

This continues to be applied to AI where people think is going to be next 12 to 18 months. Changes are coming but certainly not at the rate Zuckerberg and most people are thinking.

Re: Zuckerberg says AI agent development going slower than expected

#284

Post by mgh95 »

theturtletalks wrote: post #277
mgh95 wrote: post #266

Earlier quoted context omitted.

Interesting. Where is this product located? If I'm looking at your profile correctly, this appears to be Openship[1], which is just a collection of starter templates from something from Nextjs and something called Keystone for each vertical. Is this what it is you are talking about? I'm very curious how much revenue this is generating. [1] https://github.com/openshiporg

Yes, it’s built on the shoulders of giants, Next.js[0] and lesser-known Keystone.js[1]. Next is a full stack framework and Keystone is a CMS built on top of Prisma and GraphQL. Keystone was created by this Australian company called Thinkmill. They have used it to help businesses build custom backend systems for more than a decade. But it needed to be deployed separately from Next and they were using emotion css for t…

First I know what next is. Its standard enough.

Second, take this for what it is: your product may not be compelling in its current form. Building it to many different markets will not make it compelling. If you had a stronger revenue, please share it. This sounds incredibly thin.

Third, dont mistake building the same thing 20 times for different verticals for bonifide software skills. When a SWE builds the thing they have built before its usually to learn a language which is the easiest part of software. There is a reason a common adage in software is "9 women cant make a baby in a month". Breath is no replacement for depth.

Re: Zuckerberg says AI agent development going slower than expected

#285

Post by ex-aws-dude »

skybrian wrote: post #160
vishalkundar wrote: post #102

The gap between "useful chatbot" and "useful agent" is way bigger than people realize. A chatbot can be wrong 10% of the time and still help you. An agent that's wrong 10% of the time is sending bad emails and making wrong API calls with no one checking.

I see this as the gap between an general-purpose agent and a coding agent. A coding agent can imagine something to be true, test it, discover that it's wrong, and recover. But if you go beyond what can be tested easily, asking the agent to do real work rather than writing a patch, imagining things to be true is a problem.

It’s an age old control systems problem: open loop vs closed loop

Re: Zuckerberg says AI agent development going slower than expected

#287

Post by theturtletalks »

byzantinegene wrote: post #278
mgh95 wrote: post #266

Earlier quoted context omitted.

Interesting. Where is this product located? If I'm looking at your profile correctly, this appears to be Openship[1], which is just a collection of starter templates from something from Nextjs and something called Keystone for each vertical. Is this what it is you are talking about? I'm very curious how much revenue this is generating. [1] https://github.com/openshiporg

very much seems like a product built by an someone with no product research. The UX makes no sense for e-commerce.

That’s the neat part, you can use it headlessly. The built in dashboard and storefront are using the GraphQL API but you can deploy your own external dashboard and storefront using the same API.

We’re also very bullish that the chat interface is the universal UX now. Instead of sifting through the dashboard to change a product price or sell in a new region, you can use the built-in agent and just tell it to do that. Every Openfront comes with an MCP server that interfaces with the API so the agent can literally do anything you can do using the dashboard and API. This is where an agent running the business autonomously comes in.

And even then if you’re not satisfied with the backend API for each vertical, these Next apps can be forked and adapted to tightly fit your business instead of you messing with configs, you can make the app your own.

Re: Zuckerberg says AI agent development going slower than expected

#288

Post by paytonjjones »

8note wrote: post #221
paytonjjones wrote: post #205

Earlier quoted context omitted.

Research and experimentation on neural nets has been going on since the 70s (arguably much earlier even), but the lions share of capability changes has all been in the last couple years. Scale was really the unlock; the new pre and post training techniques and architectures are very cool and useful but they definitely aren't the differentiators when comparing to the previous era of NLP.

which non-transformer neural networks are matching frontier performance using compute scale?

BERT is a transformer! The unlock happened within transformers, yes, but they were not exactly super new or innovative architectures at that time. The scale was the main innovation that brought us from BERT to today's frontier.

Re: Zuckerberg says AI agent development going slower than expected

#289

Post by bwhiting2356 »

efficax wrote: post #115

I was worried this time last year that by this time this year, companies would have slashed their engineering teams down to a handful and everything would be driven by mostly autonomous agents with human guidance. But it just hasn't happened. Do I write all my code with an agent now? Yes. Can you just give an agent a desired outcome and let it work, unsupervised? Absolutely not. I can produce more code than I used to…

Could be lack of imagination on my part but I truly can't imaging shipping 1000's of lines of code that I can't understand (beyond low-stakes prototypes). That means there's a ceiling on productivity gains.

Re: Zuckerberg says AI agent development going slower than expected

#290

Post by Aperocky »

efficax wrote: post #115

I was worried this time last year that by this time this year, companies would have slashed their engineering teams down to a handful and everything would be driven by mostly autonomous agents with human guidance. But it just hasn't happened. Do I write all my code with an agent now? Yes. Can you just give an agent a desired outcome and let it work, unsupervised? Absolutely not. I can produce more code than I used to…

If you end up with 2 to 3 times more code. That is HORRIBLE, because it means about 50-66% of the code is otherwise unnecessary. Those are eventually going to become unmaintainable garbage.

However, if you get 2 to 3 times the code in the interim, that's probably less than what's needed. I find myself cycle through almost 10x-20x amount of code implementations to get what I want which is actually less code, simple solution and desired behavior.

Given a specific behavior, there are usually just 1 simplest implementation, whether done by human or AI. However, there are 100 ways to do it with more complexity and either handwritten or AI slop, it will mean pain down the line. We used to have a lot of handwritten complexity because of certain design pattern culture, but they used to be contained because the ability to generate them is costly. Now it's much more risky and therefore more important to have simplicity as the guiding principle in ALL projects.

Post reply on HN