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

Searching for the Agentic IDE

twitter.com

Post reply on HN
Print view View original

31–39 of 39 posts

Re: Searching for the Agentic IDE

#31

Post by pjm331 »

verial-lab wrote: post #28

I appreciate Karpathy for thinking out loud like this. We all feel the shift toward orchestration... just haven't seen a UI that fits yet. I love seeing people experiment with RTS game UIs as agent orchestration interfaces. Mostly demos so far, but there is a ton of creative potential for orchestration UIs. The biggest challenge is as LLM costs drop dramatically each year, the number of agents able to be orchestrated…

The closest thing I have found that fits right now is just Linear or [insert your own project management tool with a good api here] and then you manage agents in many of the same ways we’ve been managing human engineers for the past decade - assign them issues

Re: Searching for the Agentic IDE

#32

Post by pillsburycat »

dominotw wrote: post #29
pillsburycat wrote: post #15

I've been working on something along those lines (multi-agent orchestration IDE) for the last few months as a personal project. There are also a lot of projects out there approaching this from many different angles. Curious what features people would like to see in an Agentic IDE? Would you like to instruct multiple agents in real time (like vibe coding on steroids) or dispatch autonomous agents to solve a long-runni…

i just dont want to sit at my desk, glued to a screen/phone use keyboard/voice for 8 hrs a day . i want it to be "multimodal"

So what would you think of something that works mostly autonomously but pings you on your phone if it needs input?

Re: Searching for the Agentic IDE

#33

Post by naveen99 »

hungryhobbit wrote: post #8

I don't understand how random thoughts on X are front-page news on Hacker News. If some tech CEO makes a major announcement on X, it's newsworthy and belongs here. Anything else that's actual news is also fair game ... but all other X posts do not belong here!

Karpathy is a ceo of his startup.

Re: Searching for the Agentic IDE

#34

Post by Terretta »

tmp10423288442 wrote: post #17
hungryhobbit wrote: post #8

I don't understand how random thoughts on X are front-page news on Hacker News. If some tech CEO makes a major announcement on X, it's newsworthy and belongs here. Anything else that's actual news is also fair game ... but all other X posts do not belong here!

Karpathy is a notable researcher and broader AI leader. Among many, many other things, he invented the term "vibecoding". He also recently posted his autoresearcher project, which is using a swarm of agents to optimize the LLM training and recently produced a training process that is the fastest to achieve GPT-2-level performance using a very small model.

Sounds as though one to follow on Twitter… and not duplicate here?

Re: Searching for the Agentic IDE

#35

Post by rcarmo »

Somehow, I feel like I’m converging onto this. I built https://github.com/rcarmo/webterm as a quick solution to the “multiple agents” problem and https://github.com/rcarmo/piclaw started out as a mobile UX for a simpler, tighter “claw” and has been slowly morphing into an IDE (I just shipped UI extension support and have refactored the CodeMirror editor into an extension, and I suspect a terminal is next as well as a few custom viewers/editors). I spend more time inside piclaw’s web app on my desktop than VS Code now…

Re: Searching for the Agentic IDE

#36

Post by _zagj »

tmp10423288442 wrote: post #17
hungryhobbit wrote: post #8

I don't understand how random thoughts on X are front-page news on Hacker News. If some tech CEO makes a major announcement on X, it's newsworthy and belongs here. Anything else that's actual news is also fair game ... but all other X posts do not belong here!

Karpathy is a notable researcher and broader AI leader. Among many, many other things, he invented the term "vibecoding". He also recently posted his autoresearcher project, which is using a swarm of agents to optimize the LLM training and recently produced a training process that is the fastest to achieve GPT-2-level performance using a very small model.

> Among many, many other things, he invented the term "vibecoding".

Yeah, that's a great reason to hate him, but the person you're responding to asked why his Twitter braindroppings belong on the front page.

It should be stated, again, that Karpathy completely missed the boat on LLMs, leaving OpenAI before they developed ChatGPT, and that he convinced Tesla to pursue a visual-only, no LIDR approach to FSD that doesn't work and probably won't ever work until after LIDR-based systems have already solved FSD.

Karpathy is the AI-equivalent of Sam Altman, who, for whatever reason, only fails upward. I think many HNers like him because he reminds them of themselves. Look at this bullshit and tell me it doesn't read like something the average HNer would write: http://karpathy.github.io/2020/06/11/biohacking-lite/

Re: Searching for the Agentic IDE

#37

Post by samrus »

hungryhobbit wrote: post #8

I don't understand how random thoughts on X are front-page news on Hacker News. If some tech CEO makes a major announcement on X, it's newsworthy and belongs here. Anything else that's actual news is also fair game ... but all other X posts do not belong here!

Karpathy is more of a thought leader than most CEOs tbf. I find that instead of hyping up things to male more money, he genuinely discusses things he thinks are interesting and offers very good insights on them. He coined the term vibecoding. Even if your not bullish on AI, the mans worth listening to

Re: Searching for the Agentic IDE

#38

Post by samrus »

hungryhobbit wrote: post #26
tmp10423288442 wrote: post #17

Earlier quoted context omitted.

Karpathy is a notable researcher and broader AI leader. Among many, many other things, he invented the term "vibecoding". He also recently posted his autoresearcher project, which is using a swarm of agents to optimize the LLM training and recently produced a training process that is the fastest to achieve GPT-2-level performance using a very small model.

That's great, he sounds like a great guy ... when he say something newsworthy (on X or anywhere else), it might deserve to be here. But his random thoughts on X do not .

Let people decide that. This particular thought for example is interesting. Its about a shift in the UX of agentic development that i do feel is severly lacking. Its the kind of thing that seeks obvious once pointed out. People was to be partof the conversation. i think its fine if its not pa ckaged in a self important blogpost, its still worth reading
Post reply on HN