matheusmoreira

Anthropic has a separate meter for Fable. I used to get like five Fable sessions per week and that's it.

OpenAI has no such nonsense. No separate meter. No five hour limits. I get to use Astra at max effort on literally every task if I want to, and even this somehow lasts me several days.

Anthropic got caught playing stupid "20x refers to the 5h limit" word games with their customers. Meanwhile, I have statistically verified that OpenAI Pro 20x = 4 * Pro 5x = 20 * Plus, exactly as advertised.

I quantified cybersecurity lockouts on my code review benchmark and they were significantly lower on OpenAI:

https://www.matheusmoreira.com/articles/code-reviewing-lone-...

My benchmark also suggests even OpenAI's Sol models can match Fable performance at a fraction of the cost.

OpenAI also used to have a ton of very nice features: unlimited chat separate from codex, allowing turns to finish even at 0% usage remaining. Sadly these got removed after abuse.

As a former Anthropic customer, OpenAI is simply the better company. There is no way around it. Good place to be while the chinese open weights models catch up. Claude is good but it doesn't make up for Anthropic's shenanigans.

My code review benchmark put Sol 5.6 on the same performance tier as Fable 5.

https://www.matheusmoreira.com/articles/code-reviewing-lone-...

I understand the hostility against AI itself, but people are reacting with prejudice against AI users. People are building blacklists of "AI boosters" and their "tainted" projects. This is not fine.

Free and open source software communities were not always friendly, even at their peak... And now there are far too many projects that are outright hostile towards anyone who uses AI for any purpose. Very discouraging.

Anthropic is straight up scamming its users at this point.

I've made this argument here on Hacker News dozens of times over many years now.

These "safeguards" are actively contributing to computer insecurity at this point.

It's because they are monopolists. Copyrights are monopolies, and copyright holders are monopolists.

Abolish copyright and I guarantee they will completely change their attitude.

It literally never fails.

If you pay to avoid ads, you are merely letting them know that you have disposable income to spend on this sort of stuff. You're doing their job for them by segmenting yourself into the upper echelons of the market.

At some point, some shareholder value maximizing CEO is going to show up and notice how much money he's leaving on the table by not advertising to all of those people full of disposable income.

In this experiment, only some firms lost access to foreign labor.

The outcome could have been different had all of them lost access due to a foreign labor ban.

> Blocking archive is much more concerning.

Yes. That is an example of the tyranny I speak of.

It's a politico-technological arms race. They make their laws. We make technology that completely nullifies their laws. They need to increase their tyranny in order to enjoy the same level of control they had before. The end state is either a totalitarian government or an uncontrollable population.

Getting around these silly blocks feels good, but these governments really don't like it when we do that. I used to think that we'd find some kind of equilibrium along the way, that we'd eventually discover the government's limits: some principle they refuse to break, some line they refuse to cross... But the truth is these tyrants have zero limits. They really don't like losing their power over us, and will react with the most orwellian mesures imaginable, and they'll tell us it's all justified.

But would the corporarions raise salaries and improve conditions if they had literally no other choice? That is the question.

Unionization demonstrably leverages benefits out of corporations, so I'm inclined to think that immigration bans would have the same effect.

> You will soon have your God, and you will make it with your own hands.

Morpheus, Deus Ex.

Anthropic rewrote the MIT licensed Bun. The capability has clearly been demonstrated. As far as I'm concerned, the writing is on the wall.

And it's not just free software either. AI can quickly reverse engineer compiled software too. It's almost scary how it made short work of all the firmwares running on my laptop. Even the days of compiled proprietary software are numbered since AI can trivially make free software out of them.

We live in interesting times. The very act of publishing software at all is now questionable.

Isn't that the whole point of crowdfunding platforms like kickstarter?

Developer presents the idea and how much it costs. If enough people pitch in, then the idea is funded, the developer gets the money as a lump sum and can turn the idea into reality.

Why did this model fail?

Indeed. It's either AGPLv3 or all rights reserved, and I'm increasingly in the proprietary camp.

There are two free software business models:

1. Choosing the most copyleft, most business hostile license imaginable (AGPLv3) and charging for permission to violate it.

2. Relying on crowd funded patronage via platforms like Patreon, GitHub Sponsors, Kickstarter.

AI's ability to rewrite entire codebases is quickly killing 1, and even the biggest projects which are most successful at 2 don't make enough to cover even a single full time developer's salary.

All of my new projects will be private by default from now on.

The regurgitated approximation can matter just as much depending on what results it accomplishes. If I somehow make money out of this, then it's mission accomplished. Arguably, they might even matter more than my artisanal hand crafted software projects which never paid me a living wage.

Doubt.

If I vibecode a video game and manage to sell it on Steam, the information definitely mattered even though I didn't understand any of it.

There are drugs that nobody truly understands how they work and they are being used by professionals in actual treatments literally right now. We know purely statistical facts like "if drug X is used for condition Y it will help Z% of patients" and can only speculate as to their mechanism of action. They are used anyway and still benefit a lot of people.

Thanks, I was wondering about the exceptions!

The other OCRemix albums are highly recommended too, especially the early ones.

> NES Mario can be corrupted to execute arbitrary code in TAS

Pokémon, too!

https://tasvideos.org/3358M

https://youtu.be/zZCqoHHtovQ

Isn't GrapheneOS rootable?

And if you're not a citizen?

Right now it seems GrapheneOS is the only OS capable of resisting Cellebrite.

> suspicious screenshots were found I made 2 years ago about the Iran/Israel war

> I was barred from the US for life

Getting banned for life due to wrongthink screenshots is pretty wild. No idea why screenshots would be suspicious.

> I can understand that security don't want to take any chances with people carrying supposed explosive powders on an airport, so the system was doing it's protective job.

I'd understand rejecting your entry on that day but why would they make you persona non grata forever? That seems completely out of proportion.

I wonder how low Google will sink next time.

Respect for the GrapheneOS for pushing through regardless, even if they have to reverse engineer stuff. Can't wait to buy their phone.

There are two injustices here. The first injustice is the sheer existence of this notion of intellectual property. The second injustice stems from the selective application of the associated laws.

Intellectual property should not exist, but since it does, one can and should expect it to be applied consistently in all cases. Just look at what they did to Kim Dotcom, he basically got destroyed. SWAT rappelled down helicopters and descended upon this guy for the egregious crime of copying bits. That's exactly what should happen to Anthropic, OpenAI, Google and Meta. Anything less than that is an insult to previous enforcement victims.

Inconsistency robs authorities of all credibility. Selective enforcement exposes them for the arbitrary tyrants they are and delegitimizes their decisions.

IP could be real and LLM training could be considered fair use. IP protects published information, so even abolishing all IP would not help us since the LLM weights are secret.

Nothing short of a global revolution can fix this. Proprietary LLMs should be illegal. Either we achieve post scarcity within this generation or it's literally over.

Not everything needs to be general purpose. Allocation can be as easy as bumping a pointer, and it's hard to beat that.