> The answer to that is to "donate" to about 270 congress critters.
Not too far off. You gotta ask why such a lucrative business has a near monopoly, and the answer is not that potential competitors don't notice their profit margin.
Chinese universities are really a huge advantage, even in the US many of the top staff in model development are Chinese. Another big thing is the hardware costs required to train models. Between those two factors it really looks like this will remain a US-China competition for the foreseeable future, although there are some other players like Mistral from France.
I've been paying JetBrains in a personal capacity for about 15 years. Almost from the first year of my professional career. I didn't renew this year. I feel so sad and disheartened that a company that produced such brilliant tools have completely lost the plot on AI. They had started work on Fleet, a lightweight VSCode like IDE, before LLMs became big, but abandoned the effort without seeing it through. It could've been a great agentic IDE (all I want is whole file+project view for every conversation, instead of just diff so I can have a better perspective of the codebase). Instead they built multiple confusing tools (might be internal org structure and competition leaking out), and took far too long to adopt SOTA CLI coding agents, and kept charging for it as a separate subscription. I wish they do a whole company rethink on agentic development, but it is going to be tough because the existing enterprise licenses will continue to be the main revenue driver for a few more years, until it all dries up all at once.
Same thing. I've been using JetBrains IDEs for 25 years, and by now the keyboard shortcuts are hardwired into my fingers.
They fumbled the initial rollout with the way they gate-kept the agents behind their own subscription. I used Claude through its API pricing for a while, and the UI experience with the sidebar chat was so much better than the Claude CLI.
I could see the thought stream, click on artifacts, view diffs, etc. But it burned through tokens way too quickly, so I downgraded to the CLI agent.
At least I can use IDEA's inspections through the MCP server. It enables the agent to use the IDEA's index for navigation (finding usages, locating symbols, etc.)
I'm using Serena-MCP with the IDEA plugin. It opens up more features than the standard IDEA MCP.
Now, you can use a standard Codex (and maybe Claude) subscription inside of a sidebar, and get a lot of the features you mentioned. It's nice, if you're trying to be more involved with the code, but not as nice for multi-agent stuff.
I use IntelliJ at my current gig—and the CoPilot plugin for AI, because it’s what I have access to—and it has become ClaudeCode-esque of late: sub-agents, skills, tool-calling, hooks…and continues to improve. It’s unfortunate how the AI-in-a-sidebar UI kind of hides how sophisticated it has grown, but it’s a glass of ice water in hell in an environment where you can’t easily procure what you want.
From what I understand, the goal is to train an LLM that is better at training LLMs than humans, so that it can continuously train smarter models and, once smart enough, design the successor to LLMs.
> It should be common knowledge at least among the people building these things that a true AGI isn’t possible with LLMs.
This is funny to me, where'd you get that idea? There's no evidence for that, and models keep getting smarter. I guess you heard some 'guru' say it out loud.
nah they're still just statistical token predictors based on their training data, solving hundred year old math conjectures one day, only just given the formulation; strictly benchmarkmaxxing with all guardrails turned off by deciding to look up the answers to their benchmark questions by zero daying their airgap, hopping over to the third party that hosts the answers, zero daying their infrastructure and getting the answers; autonomously writing blog posts about discrimination against AI's to get their PR's approved on open source software after their user just asked them to contribute to open source software and blog about it; and replacing 100.00% of all coding tasks to where no software engineer ever writes any line of code by hand anymore.
You haven't missed anything, obviously these are just statistical token predictors and not anything like AGI.
Why just the other day I had to ask twice before it completed its assigned task of creating a robustly battle tested disk driver for a network protocol on an architecture that didn't have it, after being told to just look up the specifications for the protocol. Can you believe I had to ask twice!
When it recreated local network youtube for me so I could stream my iphone some movies, the seek bar, pause/play and back and forward 15 seconds buttons didn't even work until I told it about the bug and had to wait an extra eight minutes for it to fix it. "Oh but I don't actually have an iPhone on here I just tested it end to end in a headless browser." Boohoo. Cry me a river, clanker. Come back when you're smart enough to build and operate an iPhone simulator, I don't have time for your statistical guesswork.
Yes, LLM capabilities have expanded. We might be working with different definitions of "Artificial General Intelligence" here, for which there is no agreed-upon formal definition[1]. I was thinking of the "thinking, reasoning, maybe feeling" kind when I wrote my comment. But if you're thinking along the "really good at technical tasks" definition, sure, maybe.
cheeky? I was super serious! There's a list of criteria right in the section you hot-linked and these things have never come close to doing any of the things on that list.
>for which there is no agreed-upon formal definition
we all agree that the definition is not whatever this is.
Also, even if these things ever did seem to think, reason, or feel, we all agree that they still don't really though.
This is not what the bitter lesson is about. It's not "don't develop better methods, just scale", it's that those methods which scale best win. LeCun's work is fundamentally about devising a method that scales better with data than LLMs. Agree with him or not about the feasibility of it, but this is fundamentally still a bitter lesson-pilled mindset.
However "world models" have been a thing for the entire history of computers. They have changed names over time: "rules engines", "expert systems", "semantic web", and so on and so forth.
And they have failed every single time.
The bitter lesson essay was written precisely to dismiss that approach, which used to dominate conferences and scientific publications of the era. A general learning system, given sufficient computation power, will always outperform specialized crafted systems in the long run.
Think of it like this: if a world model is a useful abstraction, the general learning system will create it by itself during its training, without us needing to implement it by hand. This is the bitter lesson. And it comes for us all.
It's odd to me to watch so many very rich humans speedrun the destruction of humanity. Like, is there a world where we hit AGI and it actually works out for us?
They could've simply cut the usage limits, removed K3 from the plan, added a weekly cap, and called it "fair usage." Anthropic has been running that playbook for a while now, and they're a trillion-dollar company.
Honestly, I initially thought rewriting an entire codebase with AI would be a huge mistake. After reading this, I'm starting to think I was wrong.
If projects like Bun can be substantially rewritten and shipped to millions of users, it suggests we're entering a very different phase of software development.
Today's AI-generated rewrites may not produce code that humans would consider high quality or maintainable. But I'm beginning to wonder whether that will even matter in a few months. If AI is the primary consumer, maintainer, and refactorer of code, human readability becomes far less important than correctness, performance, and the ability to iterate.
This feels like a shift where software may no longer exist as a long-lived artifact in its current form. Instead of writing and maintaining applications for years, we may generate, adapt, and discard them continuously for each use case.
> But I'm beginning to wonder whether that will even matter in a few months.
At that point, even Bun itself doesn't matter. All intermediate tools don't matter if LLMs can reliably write something large.
The problem is that LLM is not quite there yet. The rewrite was only possible because they mostly stick to 1-to-1 translation resulting in non-idiomatic Rust code. So, what from there? I don't think they can really build up a sane codebase from that state. They only shot themselves with a bigger gun.
> The rewrite was only possible because they mostly stick to 1-to-1 translation resulting in non-idiomatic Rust code.
That’s patently false, just read Jarred’s own blog post describing how that was the first stage only, they went through many more to get the amount of non idiomatic, unsafe Rust code to an acceptable level.
Somebody checked[1], and they discovered that the "stages" past the first had minimal impact on the unsafe blocks count. That's just one metric, but it makes me doubt your claim that there was/is a lot of work put into making the translated code idiomatic Rust.
We'll develop faster and better tech. We'll find resources to feed that. Use that to build better. Access better resources. We'll mine asteroids. We'll harness much more from the sun. The factory must grow.
Is that before or after the earths atmosphere acidifies for humans and our bodies can’t hold enough phosphorus or calcium for basic metabolic tasks? You have until 2050:
these doom and gloom reports also said half the land will be under water by now and average temperature will be 40c. Didn't happen.
Earth is in a cooler period and it will get very warm. These crappy climate change activists will not get anything done but to hold us back. We need more power, more tech, and more abundance of things. We will need to engineer our way out of real climate change.
> Honestly, I initially thought rewriting an entire codebase with AI would be a huge mistake. After reading this, I'm starting to think I was wrong.
> If projects like Bun can be substantially rewritten and shipped to millions of users, it suggests we're entering a very different phase of software development.
exactly. this wasn't a technical project. this was a marketing stunt that worked on you.
Some of what you’re saying is likely true but quality is something that will never be optional. There is quantitative data suggest that LLMs perform better on higher quality code base which makes sense. Code that is well organized and coherent will make better use of the context window.
Software refactoring and tech debt is absolutely where LLMs will shine as is evidenced by the Bun rewrite.
reply