roncesvalles

>SSR but in the browser

We've come full circle.

My completely conjectural theory is that natural language is just a shitty medium for communicating knowledge and the idioms and motifs that the LLM uses are an emergent "code" or structured syntax that work better for communicating ideas.

If you think about it, "load-bearing" is a pretty commonly used concept in pedagogical writing. You could say "most important", but it's not quite the same in meaning. English just doesn't have a better word to describe a concept that occurs this frequently. The LLM's catchphrases reveal blindspots in the English language itself.

I don't understand why one would let a router handle their LLM request (other than for routing it to the cheapest inference provider for open models).

When I build an application that uses LLMs, it's tuned and tested for a specific model only. LLMs are not really drop-in replacements of each other, even different versions of the same series.

The pertinent question isn't why OpenRouter but why a router company is worth 7B.

My understanding is that there is no arrival time SLA when you buy a plane ticket (barring the compensation ladder set by EU regulators). All they owe you is getting you and your stuff to the destination.

The problem is that most of the volume doesn't come from proprietary products, it comes from API use which has no stickiness.

Claude already has a killer product (claude.ai/chat is a Swiss army knife) but just relying on people typing stuff into chat is not enough to sustain the company.

The other strategy is entrenching yourself as the LLM of choice into existing products (like ChatGPT is on Apple products).

Then you should consider emigrating to a country with universal healthcare such as Canada.

This is broken window fallacy. The $1T not wasted on healthcare inefficiencies would be spent/invested on other things, creating jobs there. Yes, there would be churn because a good number of people involved in the bureaucracy of private health insurance would lose their job and possibly their career. But these things get smoothed out. Social supports (should) exist to dampen the effects of such churn and keep the economy agile.

Imagine what it's like without that line.

Bad engineers simply don't care about good engineering. They aren't even trying. LLMs have made these engineers super "productive".

It is highly unlikely that the US is going to collapse any time in our lifetimes. Or even those of our children.

Also, let's not forget that the Trump admin is a problem of 2.5 years at most.

Trump has greatly stressed the safeguards in our system, but I don't think that he has broken them. That is actually an optimistic instead of a pessimistic signal -- an anti-statesman even in the highest office couldn't meaningfully damage the country.

or basically anybody who has an opinion about vibecoding but isn't actually a programmer

The number of times I've heard "SaaS is dead" irl and flustered responses when I asked "what does that mean?" is comically too high.

Because a non-crap product wouldn't be worth $26B. What you perceive as crappiness is the sauce that keeps you coming back.

Nuclear war is hardly an existential risk.

An all-out global nuclear war (all nuclear nations landing 10% of their warheads in populated areas) would be a disruption comparable to COVID at best.

Exactly, this is all a symptom of underinvestment in grid capacity, and datacenter companies having to cobble together whatever electricity source they could get their hands on at the last minute.

It's a complete failure of the state in having any kind of foresight.

China, in comparison, has around 2X excess electric production capacity in their grid. That's what made mining Bitcoin, and now ultra-cheap LLM inference, economically viable.

"Winner-takes-all" in the software context is never meant literally. It's more like, an oligarchy of a few winners will emerge out of hundreds of contenders.

It's not like restaurants where for example you can open a "Larry's Pizzeria" and still build a valuable business despite a Domino's nearby.

Most vertical software is just horizontal software glued together

Since software is still a winner-takes-all market, the mass-production property of software doesn't really matter.

In such markets, what you produce is either worth nothing or worth millions of dollars. For as long as it's the case that well-constructed code (with or without LLM help) is more likely to be in the latter category, the economics of software don't really change.

Even before LLMs, you could've commissioned a half-assed clone of any app you wanted from a 3rd world consultancy for a few thousand dollars. LLMs are basically Bangalore-as-API.

Yes

How is this different from hosting infra or just SaaS in general?

TFA contends that there is some fundamental shift in the economics of software, but it doesn't look to be very different from either a new SaaS dependency or racking new hardware.

>Users began expecting something fundamentally different from software: not just tools that store and retrieve, but products that reason, generate, and respond.

Not really.

>Every inference call costs money.

Not really, either. If you buy your own GPU, rack it, and run an open model, there is no unit cost. This is just expensive hosting infra. You also pay unit costs for SaaS that your software uses (things like SMS etc).

And what about the terrorism that exists today?

Many cancers are "cured" in the sense that they're detectable and treatable. My grandmother lived 35 years longer than she would have by detecting breast cancer early and getting a mastectomy.

There is currently a $800 SOTA blood test that can detect most cancers before any symptoms. Maybe a decade until it's a routine part of your annual blood test?

Whole genome sequencing costed $2.7 billion in 2003. You can get it done today using a mailed kit for $400.

HIV went from death sentence to all-but-cured in 50 years.

This will remove a non-insignificant chunk of Tor network capacity. Many of those OCI free-tier instances were set and forget Tor relay nodes, especially since they're de facto unmetered.

Nukes don't "destroy the planet", they're just really powerful bombs. Nuclear munitions have been detonated 2058 times already.

https://www.atomicarchive.com/almanac/test-sites/testing-chr...

And the best part is, I don't spend more than $20 a month on LLMs. Going manual and constantly branching keeps the contexts super lean.

Also, the correct way to LLM is to constantly trial-and-error in new/branched contexts.

Remember the LLM is not a human employee. You don't have to say "yes and" to whatever crap they produced so as to not hurt their feelings or infringe upon their creative autonomy, nor do you have to defend the correctness of your original instructions so that they don't think less of you for asking them to chase the wrong goose.

I probably generate 20-50 lines of code for every 1 line that I keep.

This is also why I think harnesses and things like Claude Code and OpenCode are false efficiency. The only way I can maintain my pace of branched trial-and-error is by using claude.ai/chat and manually extricating code fragments to and from my codebase. The human is still the best harness for production-level code.

>At Ashby, your compensation is market-rate indexed to your level of performance.

FYI, "market rate" refers to the highest compensation that someone would be willing to offer a person with that profile. When you sell something, you sell it to the highest bidder. You don't sell it to the median buy offer.

>We target your total compensation (salary + equity-to-be-vested) to be greater than the 75th percentile

So, below market rate?

>We've seen plenty of examples of apps successfully vibe coded by non technical people, including apps making real revenue.

Have we? Or is this just something that people say now, without citation?

Actually, you are right. We are working with different definitions of quality of life.

The QoL improvement from LLM is somewhere in the ballpark of going from dumb phones to smart phones for an average person (able-bodied, local to the area etc). It's nowhere near Haber-process or penicillin.