raghavtoshniwal

There was o4-mini and 4o-mini

I agree. Do you think LLMs can now handle relatively larger codebases than they did before? If yes, do you think this trend will stop at some point?

Haha - cool that you made a throwaway insult account just for this. I meant punch cards as a placeholder for programming in a bygone era, about how it feels so different, distant and detached, and that I don't relate to it.

I can see a plausible future where if we go down this route, what I call coding right now will feel the same.

I feel a certain way when I hear about older programmers who used to program using punch cards, I guess everyone in the future will think about us in the same way?

I think you're correct but your bar is too high, I think this app would be useful even if it was a lossy approximation Hemingway from his writings. As a thought experiment - I would value what a PhD who dedicated her career to studying one author and their works to tell me what she thinks about a piece of writing from that author's lens. (It's not too far from it)

> Anything else is like trying to eat a picture of a sandwich to satisfy your hunger.

I think it's more akin to you trying to recreate a different sandwich after reading a couple of their cook-books.

Isn't that a bit like saying that if common cold is solved, you will believe that human intelligence is possible? Why isn't it super intelligent if there are security flaws in the system. [Not taking a stance on wether super intelligence is possible but your condition seems a bit arbitrary]

I feel insane cognitive dissonance when I read a comment like this. I know/hope what you're saying is in good faith and you aren't trolling. Yet my own experience on how good these models have become and how rapidly they're improving makes me feel like we're talking about 2 completely different things.

Screw the benchmarks, it feels insane how much utility these models already provide in my life and that they keep getting better. I guess all my problems are simple and and "lie in the span of text on the internet", but they're still extremely valuable to me.

>subpar scores at benchmarks like SWE-bench

The last few models have remarkably improved on SWE-bench too. o3 scores 73%, this number was in the low teens 16 months ago. Willing to wager that SWE benchmark gets saturated before the end of 2025.

> aren't particularly representative of what a real coding job

I don't know about that, large swath of "real world" coding is writing plumbing and UIs for CRUD apps, they're getting really good at that as well. Anecdotally, engineers I know have gotten insanely productive with tools like Cursor.

The article you linked is from Sept '24 and points to the ARC-AGI test as "evidence" that we're not getting close.

We're in Feb '25. ARC-AGI (at least the version they're referencing) already has been solved by AI at above average human level.

>everything points to incremental progress with signs of diminishing returns.

Seems like everything in just Dec '24/ Jan '25 points the other way. These models are already helping PhDs in novel research, they're already getting super human at coding (yes yes, they're not perfect and I'm sure someone on HN has this weird coding job that AI can't replace yet and they're very excited to shit over AI), but they've already replaced a lot of real software dev jobs.

Also aren't you contradicting yourself?

> everything points to incremental progress with signs of diminishing returns

> corporation replacing employees with AI

If we have incremental progress, how are corporations going to replace employees with AI?

It feels like every major lab is saying the same thing:

https://darioamodei.com/machines-of-loving-grace https://www.wsj.com/video/events/the-race-for-true-ai-at-goo...

Even folks _leaving_ OpenAI, who have no incentive to drive hype, are saying that we're very close to AGI. https://x.com/sjgadler/status/1883928200029602236

Even folks like Yoshua Benjio, Hinton are saying we're close to it. The models keep getting better at an exponential.

How much evidence does one need to dispel the "this is OpenAI/sama hype" argument?

Forget singularity, do you think the robotics problem is genuinely hard enough that if we devoted significant amount of intelligence to it, it would not get solved?

>any day now Optimus will become sentient and replace home builders.

I think you're kidding about being "sentient" but it feels like they just have to get somewhat good at a very few tasks and we would be able to automate some large swath of manual labour. We don't need that many fancy tricks to get there. A lot of people are already reporting significant speed ups in Bio research, why wouldn't we see that in robotics?

The world’s richest man got handed this mandate by someone that was elected by millions of US citizens and he explicitly told them he will do this. Enough people in the US want to see this happen.

I don't know if enough people would not agree to highly tax the corporations if they're themselves out of work and need the money to survive.

Built a hardware device that sits between any computer and any printer and reads what is getting printed.

Primary use-case is to read receipt data from legacy POS systems without having to write software integrations.

Figuring out how to commercialise. Reach out if you have ideas!

It makes sense that one size fits all privacy policy wouldn’t be possible, but for most good faith actors, they collect data in same-ish patterns, and have same-ish privacy stances.

Why don’t we label and popularize them so consumers know what they’re getting into without having to parse the policy each time.

Agreed; but isn’t this on the same continuum as programming assembly -> IDE autocomplete -> LLM autocomplete? You’re still writing code, but generally adding abstractions has been net good (unsure of this opinion tbf, but that’s my hunch)

If this increases iteration speed for beginner devs and they learn about code quality post it goes into the real world, it’s not a bad bargain to strike imo.

I think we all partly learnt about code quality by having our code break things in the real world.

Adding a space after 'before:' works.

Edit: sorry, this invalidates the filter. My bad.

Sooo, short Nvidia?

Cool to see this Avikalp!

Hi Brajeshwar, How do I join the founder community you mentioned?

Doesn't twitter have its own datacenters too? But they also use GCP, not sure how they split the load

> Mysterious "tapes" would load it into one's brain like a program being loaded into a computer.

> That sort of thing is unlikely to happen anytime soon

Probably just my lived experience, but watching youtube at faster playback speed feels like downloading information. Increasing speeds at parts which I can easily grok and decreasing it when it takes time to understand the content.

Geohot has some bias to believe AGI would have to exist in meatspace, he has invested years into cracking self driving cars and dealing with the messy real world.

He does plug open pilot at the end.

pip uninstall pip also works, I found out the hard way.

With game streaming as popular as it is, don’t these players post their esports career have an existing fan base to build a career by just streaming? Something very hard to do in traditional sports.

I’ve been thinking about how NFTs are different from ordinary “costly” signals.

NFTs directly come with a price attached. An attached dollar value is the problem.

The difference between receiving an expensive ‘looking’ gift and and a gift with an expensive price tag attached.

Maybe I’m wrong, but being subtle about the effort/cost is a big part of these signalling. NFT skips all of that.

> Historically that’s never been the case

We overestimate how old history is here. There is a case to be made about how we've enjoyed exponential growth in cosumer technology over the last few decades but that could slow down on a few fronts. For ex- display resolution has reached "good enough" fidelity for a while.

I certainly hope you're right and we find cool, novel use cases but I wouldn't be certain. I personally have not thought about bandwidth for a few years now. Meanwhile I remember the speed bumps being exciting earlier. Diminishing utility is real.

If you're quick at simple calculations early on in your life, it probably has a compounding effect on the other aspects of your life. Not sure how you'd prove it but alot of people who have a Math/Science acumen, do so because they're above average early on and it gets reinforced. Mental math could give one that edge.

Tangential : Remember reading somewhere that 'lifetime' means the product's lifetime, not the user's

https://www.oslash.com/ does something similar, with a nicer interface and more collaboration.