ipsod

Still seems like the same driving force of curiosity behind both, but one wants to know and one wants to do. It's like yin and yang, like two sides of the same coin.

I'm into engineering because I want to create what's never existed.

Science is more about discovering what already exists.

if curiosity > pain avoidance: engineer

I don't get that experience at all. LLM's both relieve burden and improve speed. The only trouble I have with AI is when I run out.

That, and I miss coding... but... this new thing is its own puzzle, and I still do lots of pseudo-code and such, to talk to the AI, which kind of scratches the itch.

It's ridiculously fast, and generally very good with real-world knowledge. It's my go-to for anything non-coding, and for quick fixes/questions/prototypes.

If metrics measured what most people need from an LLM, it's competitive, IMO, and with their TPU chips, it's probably economically competitive.

I don't blame you for thinking that. I've been there.

Nah. I've committed over a million LOC in the past 4 months, discounting personal projects. LOC is a fuzzy measure that doesn't tell much, but, it's easier to mention that than to describe the details of my work.

I'm pretty confident that, presently, with AI, I get more done than I could with 5x hours, but less than I could with 20x hours, and 10's a nice round number.

I'm not defending AI, I'm attacking your logic.

A team of programmers is "worse" than their best member. Worse quality of code, worse thinking. You can usually get more done with a lot of mediocre hours of work than a few brilliant hours, though.

AI I've used isn't a better coder than I am - it's just got a lot more hours in an hour than I do.

No. RoR lets you build a certain class of web apps faster. AI lets you build almost anything faster.

That's a hot take. AI has fundamentally shifted things. Start-ups can start up 10x faster, and HN is start-up flavored.

If people quit talking about AI, it'll be because it's become foundational infrastructure, so it's not the primary focus of the articles, but just the atmosphere they all exist within, like the computer, or the internet.

Nobody knew that LLM's were an option. The architecture was basically waiting there for someone to say, "do that, but turn it up to 11," if I understand right.

Is the alternative is to give the power directly to the powerful? Not only will they hold the guns, but they decide what the markets will bear - often-enough with food for the poor becoming optional, if history has any lessons.

Which non-capitalist system does not ultimately concentrate wealth and power even further? And, if nation-states are organisms, and survival of the fittest is a law, which of those systems can survive in the real world? Excepting perhaps those propped up by a disproportionate natural wealth and homogeneous culture, where it is easier for people to continue to believe in "our good" when hard times come.

Capitalism is the only system that pretends to respect human nature - gives us a lever that says, "want more stuff? provide more value." It gives the self-interested a healthy outlet that the other systems lack. It must be managed, though, by power, and when the two collude you get fascism. So, I guess maybe I share your view, that lobbying breaks the system.

I think it might be like, a praise of inanity.

That's what I do. Both for this reason, and because I don't want to log into my personal gmail on work devices.

I don't necessarily feel safe, though, and have moved my important stuff off of gmail.

Because it says it more often than kids say "six seven".

Not only do they have access to the hardware - they've been developing it in-house for years.

Gemini Flash is the one I get most excited about, because it's so fast and so good at real-world knowledge, and it's improving so fast - look at how much the benchmarks improved in ~1 month. It's just categorically different than anything else.

Also, I use it every day, and it just got ~10% better at coding, according to the benchmarks. How is that not exciting?

I haven't had any issues lately.

They've been my bet to win the AI race for a while. I was starting to doubt, but this 3.6, 3.7, and 3.8 arc has anchored me.

3.5 pro doesn't exist yet?

IDK if it's smaller, but I know it's way faster. In one test I did, Flash 3.7 high was ~9.4x faster than Luna High.

But, also... Sol crushes Flash 3.7 at writing code in a codebase of any size beyond "tiny".

Flash is my go-to for prototyping, and basically anything that isn't writing production code.

> This is of course important for protecting you.

Well, that's very kind of them. I am constantly impressed at the kindness of our governments, and the recent growth of that kindness. I guess that, with all of the power that modern technology is giving them, they're finally getting to live out their heart's desires of being very, very kind.

Do you find the variety helps? I've migrated away from such complexity, and I simply have multiple agents of the same model run the same prompt (usually Sol 5.6 high or max), and generally this gives plenty of adversarial input. I'd be curious to know how much difference it makes to run multiple models.

You generally don't even have to convince it, or at least I don't. I just paste the error into the prompt window, and say "you got blocked, try again", and it'll just say, "Oh, that's because ...", then do it.

I guess we might actually have similar feelings, except, you say "only speed", and "only economics". Only time and only money. As if those aren't two of the most valuable currencies on earth.

> You are pretty fast at jumping at conclusions.

Yeah, because I can see some of my own thoughts and experience from a long time ago in your words, and my experience tells me loud and clear that they're outdated and wrong.

OpenAI is what I use most. Sol 5.6 still rejects a few requests a day when I'm working on web apps, but, overall, it's not too bad. I wish it'd auto-resume and try again, instead of waiting for me to intervene, but it's rare enough that it's not a huge deal.

It probably doesn't help that I'm using frameworkless PHP - I imagine a lot triggers could be avoided if I was using a framework where secure features were baked in.

Web apps are where I have this trouble.

Making a web app secure is literally just finding and patching vulnerabilities, instead of finding and exploiting them. You could have the AI "try to make this app secure", find what it patches, and use it for exploits, and the AI can't know if that's what you're trying to do or not. I don't know how you can get around this. I get around it by not using Anthropic products, at present.

Nobody convinced me AI was useful. At this point, nobody could convince me it wasn't - it's like going from whittling stuff out of wood to having a kitted-out fully-manned machine shop.

It's clear from your words you lack experience and skill in this area - you've only got opinions.