Hacker Newsnew | past | comments | ask | show | jobs | submit | dbreunig's commentslogin

> What you get is a beautiful animation that is 100% accurate and free of hallucinations

How does he know?


Wrote about this awhile ago, and it hasn’t changed:

> Time and time again, when talking to people who rely on ChatGPT, Claude, Perplexity, and other general AI tools, I hear them say, “AI is incredible. It handles nearly everything I throw at them.”

> “What does it fumble with?” I’ll ask.

> “Well, it still gets things wrong when it comes to my line of work.”

https://www.dbreunig.com/2025/04/08/on-ai-observational-comi...


It's the kind of overstated unjustified claim an LLM would make.


I thought of the "does he know" meme when I saw this.


I think these same people believe everything they see on TV, unless the thing on TV is something they already knew about.


Gell-Mann Amnesia. Recognises the failings in their own field of discourse, assumes all true about anything else, same source.


I got an early tip on this and used it to build a brand mood board and it crushed previous attempts with Claude. Highly recommend.


Elizabeth Lopatto at The Verge makes a strong case we _do_ have proof that Musk is actively gathering and throwing fuel on the fire: https://www.theverge.com/ai-artificial-intelligence/929129/s...

> But the thing is, Molo doesn’t actually have to be good at this job, because the point of this trial isn’t to win — though I’m sure Musk wouldn’t mind a win. The point is to punish Altman, Brockman, and OpenAI. Musk has done that pretty thoroughly — reinforcing in the public’s mind that Altman is a liar and a snake. This morning, I read an exclusive in The Wall Street Journal that assorted Republican AGs and the House Oversight committee wanted to look into Sam Altman’s investments. References to the trial are peppered throughout the article.


Oh sure, the trial is maybe 5% a Hail Mary and 95% about distracting and disrupting OpenAI. I read "behind this" to mean more-clandestine moves, e.g. planting stories, conducting and leaking oppo, amplifying negative media on X, et cetera.


It does seem like there is a ton of negative PR and sentiment on social media (including HN) about anything Altman and even Dario do. Like, way more than warranted. It looks more and more like a coordinated campaign, a la https://paulgraham.com/submarine.html

Elon even explicitly threatened the OpenAI guys that they would be "the most hated" people on earth, and given what we've seen him do with Twitter, I strongly suspect there indeed is a submarine with Elon at the periscope.

Altman may be getting the brunt of the AI backlash, but the impact of AI is still extremely preliminary, and it will happen regardless of anything he does. As you mentioned, it doesn't help that these guys are telling the world AI will disrupt all the jobs but... at this point, I think they're just being honest.

As shifty as Altman is, I wonder how he gets more hate than Elon, who has objectively done way more concrete damage to the world.


Everyone knows a lot of the scumbag things Elon has done/does so it's not really worth talking about until he does something novel, people are still shedding light on the scumbag things the Sam has done so naturally it's being discussed more frequently as people share what they just learnt.

It would be fair to argue that in a just world Elon would suffer more consequences for being more of a scumbag than Sam but we all know justice doesn't apply to the rich in the US (occasionally this seems untrue but only because other rich people are pissed off at the rich person and they want them tarred and feathered).


Among benchmarkers its a frequent topic. Qwen BURNS reasoning to get its scores.


Model testing and swapping is one of the surprises people really appreciate DSPy for.

You're right: prompts are overfit to models. You can't just change the provider or target and know that you're giving it a fair shake. But if you have eval data and have been using a prompt optimizer with DSPy, you can try models with the one-line change followed by rerunning the prompt optimizer.

Dropbox just published a case study where they talk about this:

> At the same time, this experiment reinforced another benefit of the approach: iteration speed. Although gemma-3-12b was ultimately too weak for our highest-quality production judge paths, DSPy allowed us to reach that conclusion quickly and with measurable evidence. Instead of prolonged debate or manual trial and error, we could test the model directly against our evaluation framework and make a confident decision.

https://dropbox.tech/machine-learning/optimizing-dropbox-das...


It's not just about fitting prompts to models, it's things like how web search works, how structured outputs are handled, various knobs like level of reasoning effort, etc. I don't think the DSPy approach is bad but it doesn't really solve those issues.


> it's things like how web search works,

Arguably you could be running those tools on your end

> various knobs like level of reasoning effort

This can be subject to optimization too


funnily enough the model switching is mostly thanks to litellm which dspy wraps around.


No reason it can't. I know people currently generating specs from existing code; just gotta write the pipeline.


Last year they pushed out an update stating if any “Meta AI” is left on, they can access image data for training,

I turned the AI off and used them as headphones and taking videos while biking. After a couple rides, I couldn’t bring myself to put them on because people started to recognize them and I realized I didn’t want to be associated with them (people are right to assume Meta has access to what they see).

Meta Ray Bans, if kept simple, could have been a great product. They ruined them.


I think public shaming of that spyware should be a social norm.


Check out “Recursive Language Models”, or RLMs.

I believe this method works well because it turns a long context problem (hard for LLMs) into a coding and reasoning problem (much better!). You’re leveraging the last 18 months of coding RL by changing you scaffold.


This seems really weird to me. Isn't that just using LLMs in a specific way? Why come up with a new name "RLM" instead of saying "LLM"? Nothing changes about the model.


"Think step by step," was just a sentence you appended to your prompt.

It ended up kicking off reasoning training which enabled the massive gains in coding, tool use, and more over the last 18 months.

So yeah, it's "just using LLMs in a specific way."


RLMs are a new architecture, but you can mimic an RLM by providing the context through a tool, yes


New architecture to building agent, but not the model itself. You still have LLMs, but you kinda give this new agentic loop with a REPL environment where the LLM can try to solve the problem more programmatically.


Author of the post here.

I didn’t say AI was bad and I acknowledged the benefits of Electron and why it makes sense to choose it.

With 64gb of RAM on my Mac Studio, Claude desktop is still slow! Good Electron apps exist, it’s just an interesting note give recent spec driven development discussion.


Not coming at you at all, AI is a touchy subject on HN nowadays in any capacity and brings out the worst here.


I keep saying this, it’s my new favorite metaphor.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: