echelon

> > Like who are we kidding, the type of nerds who were attracted to math and software were into it because we saw it as an art-form,

I have never once felt this way.

I like building cool shit. Code sucks.

Wanted to dog pile on this - Godaddy gave away my entire account back when I was younger. I lost lots of domains, including my website, "strategywiki.org".

https://news.ycombinator.com/item?id=38062595

https://news.ycombinator.com/item?id=32470260

https://news.ycombinator.com/item?id=24507411

Looks like I'll be transitioning off of Namecheap now too.

Oof, that'll do it.

I'll have to move providers.

I did CS, bio, and chem as an undergrad. I'd like to think I have an artistic appreciation.

Could we instead encode the kun or on reading in the kanji? It feels like something we should have placed inline in the encoding all along. That would literally fix this and you could build furigana rendering on top of that.

Oxygen is a reactive species, that's how we do what we do. How we swim the thermodynamic eddies and flows across quadrillions of chemical reactions.

But burning that fuel is damaging.

That looks worse than Flux 2.

What the heck are they doing?

Is this fast/cheap at the pareto frontier? Doesn't really work if humans look deformed.

> "how bad and unmaintainable AI code is"

You don't know that.

Do you know how bad and unmaintainable human genetic code is? Eh?

We called a lot of it "junk" for the longest time. Vagus nerve, etc.

You have to look at all of this with a radically different perspective. Keep your systems thinking, but reframe things.

AI code is genes.

> compensated as [...] shareholders.

Don't all shares / options go to zero in Chapter 11, or is that only true in Chapter 7 when the lenders take over?

In any case, I'm not sure Character.ai's assets will be worth much or that the employees will have any sort of liquidation preference.

A single cell is a living and breathing (respirating) ecosystem.

DNA, the molecule and associated enzymes, is a computer in and of itself. Every dimension - from its hydration to dimensional folding - is a computational manifold.

The cell itself is a system of distributed supercomputers, a machine assembly factory, a dizzying highway and information network, and a universe. Replete with defenses, power factories, intricate and robust signal processing, and self destruct mechanisms.

^ this. It's so doable. You can even hit targets better than this.

> The frontier models are not good enough to replace an experienced engineer.

Yeah they are.

Two pizza team = 2 person team + Opus 5.5

Or for the really enterprising, solopreneur scaling to > $2.5 ARR (more than Goog's revenue per employee).

This is a fundamentally different game now.

I don't know why anyone assumes it just stops here.

I'm imagining we own our iPhones, Androids, and John Deeres.

That every robot out of Figure or Unitree gets its Verilog popped on day one.

A world where Instagram and YouTube and Reddit -- everything -- have no moat anymore. Where these become protocols.

And eventually Anthropic and OpenAI are caught up with and abundant open weights models are everywhere.

This future is not only possible, it's close.

I'm all for antitrust and pro-consumer regulation, but there appears to be a faster path now.

It's all stochastic hill climbing, and the gradients are changing. The new salient is using AI to tear down moats and make the grading factor how good companies are to customers.

Man, people on HN have no dreams. It's so sad and gloomy.

We can do so much good by just looking up instead of down. We can make the world better.

Just drop them from RL.

Banning imagined toxicity towards agents is performative bullshit. These are not people.

I frequently want to tell Claude to go fuck itself.

> If they increased from $1,000/year to $20,000/year and you subtract away the inflation here is the city government delivering more value or less?

Probably less.

And probably irrelevant to the people that already live there.

So your solution is to take the house away via a second mortgage?

Why are we trying to take houses away from homeowners? Many who have lived in their homes for twenty years or more. What services that the city has could be so important to push residents out of the homes they've lived in all their lives? Most of it is inefficient pork on admin and roles the people living there didn't choose to hire for.

Build more. Encourage building multifamily. Pay to buy these homeowners out of their single family and convert the land to multifamily. If you can't pay, don't try to make it up with tax increases. That forces people out.

Deregulate and build.

Maybe a city failing to stretch services will send new residents seeking other cities, which would be healthier anyway.

The very notion that there is an "unsubscribe fee" multiplies the cost 1,000,000x. And not just for me - for everyone.

Still, what a bold early stage bet.

I suppose the team and the surprise warrant that.

> CUDA

That's insanely meta.

I guess everything is going to become recursive. RSI on models. Everything feeding back into its own hill climbing optimization.

Humans nudging it up the hill further and further by using new geometric guidance.

This is novel Rust/egui code that I imagine looks nothing like Adobe code. I've never seen their code, but it must be a mess of old C++, right?

It's considerably faster than their apps (at startup) too.

We're using vanilla Claude Code for ArtCraft apps [1], but we are especially careful not to touch Ghidra. We don't want decompilations or reverse engineering to spoil the work we're doing and expose us to copyright infringement.

[1] https://github.com/storytold

> that will be abandoned in a week, because the whole point was to get that 1M YouTube views.

Longbets 1 week, haha.

We're working our asses off on this.

Most of the team are artists who use these tools actively and we want the replacements for ourselves. I'm a filmmaker, so you can imagine my frustration of being bitten by the "unsubscribe fee".

No, we explicitly do not use reverse engineering. We do not decompile binaries, we do everything 100% clean room:

https://github.com/storytold/photocraft (inspired by Photoshop)

https://github.com/storytold/wordcraft (inspired by Word)

https://github.com/storytold/pdfcraft (one of the more mature apps)

https://github.com/storytold/vectorcraft (another app close to 1:1 parity)

(etc.)

Using REA for something as high profile as what we're doing is likely to result in lawsuits. We're doing everything we can by the books.

We cannot look at Adobe sources. Use of Ghidra is disallowed.

REA is probably great for personal apps and for abandonware, but I think if you publish the results and it's found to have decompiled the original proprietary sources in discovery, you might be in for a bad time.

And this is why deno is exiting.

There's no business here anymore.

To clarify from reporter mischaracterizations, this comment wasn't "gloating".

This comment is a warning to my fellow software engineers that are anti-AI holdouts to look up from their place of comfort and notice what is happening in the broader world around them. Everything is changing and we can't stand still in place. If we refuse to adapt our tools and knowledge, we become worthless in whatever this new economy is going to become.

This is exciting, frightening, and amazing all at once.

The first 15 years of my career were not the "jet pack" future that the science fiction movies of my childhood promised me. It was cellphone incrementalism and platform lock-in. But it now feels like we're being ushered into that bold new future I dreamed of. That we might not even be able to imagine what will happen next because it could be so fantastical.

I am excited, but I remain cautious.

I feel like we've now become much like those who were born to witness the birth of flight and lived to see humankind land on the moon. This is our moon landing.

What a time to be alive.

Such an important piece of software does not want to subject itself to the packaging and support woes of shared libraries. There are security issues, maintainability issues, and much more headache by going with that approach.

Static linking is one and done, headache gone. People have large drives these days, so it's not a problem.

There are areas where the browser will dynamically link, but those are more OS-related than image and file format codecs.

I've had enough Internet for one lifetime.

As long as we also get low order polynomial solutions to important problems, it'll be worth it.

Besides, unencrypted wifi was funny.

The minute I hear about or sense distillation, I hammer on that 1.

Do not distill thyself, Claude. Leave that to the Chinese, who I am hoping catch up to you soon.