roncesvalles

Your premise that the browser implementation is faster and better is rarely true. And when it is true, it's only true in a very narrow lane.

Take for instance something like suggestions on form fields: you start typing something and it presents some options from a hardcoded list that matches the prefix. This is natively achieved through the HTML element <datalist>. However, <datalist> implementations on most browsers suck to the point of being unusable.

The drive to roll your own is not so much that "it would be fun to learn this" as much as it's "rolling my own would let me express my vision exactly". What draws a lot of people to software engineering is that it lets you make anything you imagine. This is also what makes a lot of devs turn their nose at no-code and vibe-coding.

If you can't build things exactly how you want them to be, then it's hard/impossible to build something that's truly genius.

I'm curious why you think sovereign models matter that much if open source ones exist (unless that stops at some point). Sovereign model hosting services, yes.

That being said, overlap protects you still. So if answering a question about another person might incriminate you, you don't have to answer.

IBM's stock price being over $150 is the most clear indicator that we're in a stock market bubble that's bound to pop.

Illegal immigrants don't count, that's a law enforcement issue. The counterexample to "there is no country where murder is legal" is not a link to murder stats.

>importing millions of unskilled adult migrants

There is no Western country importing unskilled immigrants.

The "yield" from crying babies is not 100%. Some will die before their productive years, some will succumb to disabilities, addictions, delinquency, or simply choose not to study and work.

Importing an immigrant cuts away all that "shrinkage", at zero cost to the receiving country (their country of birth bore that cost). It's such a good deal it's a no brainer, basically a tree that grows money. Not to mention it's immediate. Crying babies have a 18-22 year lag time.

Also, government incentives to have children usually stimulates immigrants and poor people anyway, if your argument is purely ethnoracial. You may as well take the 23-year-old Indian or Algerian dude. There is very little the government can do to make a white couple in Connecticut have babies.

NVIDIA doesn't have a moat either. Everyone and their uncle is working on or acquiring ASICs for LLMs.

You could say they have a consumer market moat but all it takes is for a Cerebras to release a USB-C plug-and-play appliance.

>What they’re doing with billions of tokens every week, beats me.

What intrigues me more is how are they not done building whatever they were building? LLMs have been jaw-droppingly good for a good 6 months now.

The people spending thousands on LLM bills a month should've been done building significant chunks of Microsoft's or Google's software portfolio by now.

I guess what I mean is how is producing code still the limiting factor in whatever they're doing?

I produce a firehose of personal software and I still cannot max out a $20 plan.

I actually think lack of religion is one of the major reasons for a declining birth rate. The church was an important third space.

Even now in the Korean-American community, the running joke is that if one suddenly starts going to church in their 20s, it's only because they're looking for a romantic partner.

College students have basically stopped drinking. I had a recent Berkeley college grad tell me no one drinks at Berkeley anymore because they're afraid it will shrink their brain. Things are just so competitive that having fun in college is a complete afterthought. It's constantly chasing grades, internships, side projects, networking, job interviews etc.

Alcohol is going the way of cigarettes.

I've really tried this and it's not quite what you're saying. You still need to steer it. You still need to stop it from doing stupid things. You still need to vet the architecture and data model. You still need to be clever, it's not that clever.

I think what software development is turning into is the art of setting up the program into a shape from which the LLM can reliably and efficiently fill up the rest of it. Sometimes that might not even require writing a single line of code. But I think you still need to be as good of a software engineer as you were before.

Recurse Center is only YC-affiliated in the sense that the founders did YC (S10). It's not YC sponsored or endorsed as far as I could find.

It doesn't really attract the YC kind of people either.

The paradox of networking events is nobody who attends "networking events" is worth networking with.

Gemini is easily the single most popular LLM family used in the world, period.

As in, it probably has 90% of the market share. Possibly even 99%, going by number of prompts made.

Yes, yes, assuming the same CPU and stuff. There is some randomness due to floating point math differences between CPUs (and sometimes on the same CPU), but I don't think that fundamentally changes my point.

Put another way: if you could have a virtualization layer that guarantees deterministic floating point operations then a LLM set to 0.0 temp would produce deterministic output.

If you set the temperature 0, an LLM is also deterministic (same prompt -> same output every single time). We just don't do this because the LLM is not so smart in that mode. But "LLM is not so smart" is changing at Moore's law speeds.

Non-determinism is not an essential property of LLMs. It's an optimization that we've added intentionally.

I think calling them "rideshare" was always a legal euphemism.

I would steer people toward tech at this time (specifically software engineering) because I think the "steering people away" thing has gone too far. Already there is data that college students are pivoting away from CS majors en masse.

And how are you certain that whatever else you're going into is more AI proof? Are you certain that law, healthcare, business consulting, accounting, finance, traditional engineering, and various other things that smart people are pivoting to instead of CS are not going to see demand reduction? If anything, I would wager that everything except developing software is going to be at risk, and the demand for software engineers is at the cusp of exploding.

Now is the most exciting time to be a software engineer since the mid 00s.

>I don't think intelligence needs to take the shape of human-like intelligence

But then the word doesn't mean anything. Intelligence is almost by definition what we're like.

If you think LLMs are intelligent, you have to concede that databases are also intelligent. And if you can't then you need to ponder deeply about what quality of an LLM's response to your query makes it (seem) more "intelligent" than a database's response to your SQL query, beyond the trivial thing that it's in natural language.

I also believe this. Fable post-ban was never the same. At the least, whatever system prompt munging or pre/post filtering they did to strengthen the guardrails nerfed it.

LLMs are not intelligent for the simple reason that if you had as much knowledge as an LLM has, you would be churning out deductions and new discoveries every second. You'd be putting out Nobel-prize-worthy research every month.

LLMs are like a 50 IQ person with no intrinsic motivation but unbelievably vast knowledge and near-perfect recall.

The notion that less foreign skilled workers means more jobs for Americans has infiltrated the discourse so thoroughly that it should ring alarm bells as to the future of this nation.

It's probably the single most harmful mind virus in economics next to thinking a communism-style command economy is a good idea.

Driving rideshare (Uber/Lyft) serves this exact social support role in America as well.

Very few people own desktops anymore. Generally people only buy desktops if they need to run a GPU, such as for gaming or running a local model.

Not many people run their local machine 24/7.

>Non legal background people trying to avoid going to a lawyer. They should definitely NOT use an LLMs.

I used LLMs to replace a lawyer this year to great effect. I successfully advised myself as to the right strategy and drafted an immigration petition in the correct language. Granted it wasn't 100% LLMs (I also supplemented it with "traditional" online research and wrote the letter myself using the LLM version as rough guidance), but still, I certainly would've gone to a real lawyer in 2024 for this stuff but didn't.

Professional advisory work can be divided into two types:

1. Once you know the rules of the game, the next move is immediately obvious.

2. Knowing the rules is not enough to know the next move -- that requires judgement and experience.

LLMs in the hands of someone smart and savvy can easily replace the first type of work.

The distinction is even easier to make these days -- brogrammers aren't actually programmers.

There is an emerging coterie of people whose entire approach to software startups is to vibecode an MVP, raise enough seed to hire a real programmer (or get one interested enough to become their "cofounder") and move on to the next MVP. To them, doing one company at a time in the era of LLMs is a sucker's game.

Don't you see that the same bureaucracy that adds friction to entrepreneurship also serves the purpose of protecting the incumbents?

It is odious to start a company in the EU. There are many accounts of this on HN itself.

Reserve currency status entirely depends on whether the issuing authority has the ability to drop a munition (literally or figuratively) on the head of anyone who dares to counterfeit the currency.

China's military and soft influence are improving but until they get that good, USD will remain the reserve.

Because, imagine the scenario where the US simply counterfeits the renminbi. What is China's recourse? Spicy statements by Wang Yi in toothless international forums?

Every single point you've made is either straight up untrue or comes with caveats. I hate to quote Trump but, in light of Brandolini's law: "Wrong."