Absolutely.

So what's the best primer we can share for getting started with local llms?

Buy a laptop with some RAM, and a decent GPU, install llama.cpp, or Ollama, or whatever you go, get a harness, you're off to the races.

You want to run an LLM on a laptop? It's going to be converted into a toaster real quick

It already enshitificated from the very begin

> Let the enshittification begin

But is there a solution to this though?

How does one stop 'enshittification' of everything exactly?

I'm not talking just about Anthropic.

One thing Doctorow suggests is to remove anti-circumvention, thus allowing adversarial interoperability.

Is that a principle that is ethical in general?

Like is it ethical to put in my ToS that you get a free api but you can’t use it to make a competing service? I feel this is quite common

Competition Enforcement

Strong privacy laws

Tech unionization

Mandated data portability.

Has these successfully combatted against enshittification entirely?

Which places has this been successful?

What are you talking about, these haven't even been tried because the corporate lobbyists know they work..

I'm just asking for successful examples so we can apply it more broadly to stop companies enshittifying enmasse.

Do you have multiple successful examples of stopping enshittification entirely?

Ah yes, citizens united, the great conspiracy theory. Yikes, I can at best assume this is a shill account here to FUD for the tasty boots.

???????????

What is wrong with asking for solutions?

Competition is important, but the way you fix that is by fixing the terms of copyright and patents.

Intellectual property, at it's simplest, is a government mandated monopoly granted to allow for progress in the sciences and "useful arts" (think trades really, not arts and crafts). All monopolies are bad, but it was only to be granted for a limited time so it was fine.

Corrupt politicians ruined the entire trade off though, and now it seems that copyrights and patents last for 100 years plus the life of the universe. During which time no one can really compete fairly.

If we want rapid progress in science, we want short patents. Instead we opted for long patents and a higher stock price. R&D budgets have been shrinking ever since.

The solution is mostly not to go public. Valve is a good example of this.

Do well enough that you don't need outside investors, or if you take on outside investors select them carefully and ensure they don't expect an IPO.

Valve is a bad example.

For game developers, they are a nearing close to a monopoly with Steam, anything PC and for distribution has to go through them and they take a cut, any other places aren't as effective.

Valve has no real competitors or competition.

Unless Valve faces intense competition to change, enshittification will be inevitable since they know nobody can really challenge them.

This is regardless of self or outside funding or no plans to IPO.

Eh they do have competition, from e.g. Epic and Battle.Net and what have you, it's just that their competition is public and universally worse off for it.

Their competition is beholden to quarterly reports, and can't do the sort of long term investments that would be needed to compete effectively with a company like Valve, which isn't beholden to quarterly reports.

Valve doesn't need to enshittify because they aren't subject to shareholders pressuring them increase the profit. As a private company that's pretty much printing money, they don't need to do anything... which has largely been their entire strategy. What would Gabe even do with even more profits, buy a third and fourth yacht?

> ...As a private company that's pretty much printing money, they don't need to do anything... which has largely been their entire strategy.

The need to do 'nothing' works if you are a monopoly or have little to no competition.

Epic and Battle.Net are not real competition.

What is super odd about this is no-one expects Anthropic to prioritizes shareholders.

except the employees will be compensated with equity, so they will have an incentive to pump the stock price to increase their compensation. I've seen this play out (and subconsciously been a part of it) for over a decade at Google. The early, super rich leaders won't be the ones to enshittify, but the next generation of leaders that are trying to get rich will feel the stock price reward signal.

Opus 5.5 has been terrible for me for the last week - so seems like they got the memo.

Luckily switching is super easy so I'm now back to codex (and hopefully carbon soon)

[delayed]

This. It will be very interesting to see how plans are going to change and how quickly. Maybe they won't and they'll stay relatively affordable. We'll see.