Hirom @ Hirom @beehaw.org

Posts

163

Comments

1,160

Joined

3 yr. ago

This AI bubble is fundamentally fuelled by too much money that can’t get sane investments — so it’s looking for insane ones.

There’s a real tech revolution going on right now! It’s renewable energy. Wind and solar and batteries are all super-cheap and getting cheaper, the tech is advancing by leaps and bounds, it’s doing amazing stuff.

But that’s a regular business that doesn’t look like a venture capital lottery ticket — so it doesn’t get hype-filled headlines in every newspaper from venture capital PR offices.

https://pivot-to-ai.com/2025/04/10/a16z-raising-fresh-20b-for-ai-bubble-investments/

The finance sector directing the world's resources to the wrong projects is scary dangerous. More than AI itself.

It's a hint that F-Droid need our support.

Elle s'est lancée dans une expédition dangereuse en connaissance de cause. On n'a plus de trace de vie. Et une personne est déjà morte en tenant de la secourir.

À ce stade arrêter les recherches est plus éthique que risque les vies d'autres secouristes.

Alan Tudyk would probably have been a fine detective Spooner.

Will Smith is a fine detective Spooner.

It's the casting's director choice, it's normal for actors who were not picked to be sore.

Could there be collaborative moderation where node admins volontarily block other nodes or onion services that host CSAM?

I wonder if that's technically possible for a hidden service directory or rendez-vous point to do that. It wouldn't completely kill those services, but may at least make less Tor resources available to them.

The trickiest question may not be technical but organisational. The Tor project probably doesn't want to be in a position to validate/decide block requests. Different entities all over the world would ask for blocks, for many different kind of contents.

The expression "Lightweight cryptography" reminds me of export ciphers. Hopefully the expression is the only similarity.

Export ciphers are unsafe algorithms used to comply with export control rules. Export ciphers introduced vulnerabilities that caused troubles decades later.

Lightweight cryptography selection by NIST was more transparent and security-focused. They accepted 57 candidates in a multi-round process. There was third party security analysis, discussion during workshops, and a report with results and references to published analysis.

Reunification in our lifetime would be nice, if that happen through diplomacy and referendum rather than force.

I don't know about a PRK-like state, but anyway my opinion shouldn't matter, it should be Koreans who decide.

F-Droid store/repo is safer than Google Play Store. Google's own store has many shady apps, including spyware and occasionnal malware.

I don't get the argument that Play Store is safer than others.

Which Debian distribution are you using, stable, testing, unstable?

I take care of a couple machines for family members. Those have Debian stable with automatic update (unattended-upgrade). I can't recall the system or packages ever breaking. At most users are a bit confused when an update change the UI a bit.

Sticking to stable and avoiding third party repos gives a pretty solid system. Only developers or sysadmins might consider Debian testing. Only people working on Debian itself should use unstable.

How soon can those help cleanup Fukushima and Tchernobyl?

I assume the high price and limitations would make them a hard sell for a typical factory. But those could be handy in places where humans can't go safely.

I agree it's stupid. It's also wired in the incentives business and their management have, and the way the economic system works.

Investors typically give their money to companies they think could grow in the short/medium term, and get them good return on investment. Managers performance reviews and bonuses depend in large part on the last period's business results, not the results 15 years from now, because that's what investors wants.

Having smart smart people in leadership who care about public health and longevity will not (be enough to) solve this. Even they are regularly pressured for short term results, and may loose their job/investors/customers if they don't produce those results. Improving regulations and incentives to account for longevity and public heath is necessary.

Yes, no tanks.

Everyone, please vote and voice your support for CBAMs, ie Carbon Border Adjustment Mechanisms.

This is the polluters pay principle applied to imports, in the form of carbon tariff. It can be very efficient is enough countries put it in place and hold the line.

They seek to maximize quarterly revenue, their next quarter in particular. Not in 15-30 years.

Don't expect a company to decide on its own to prioritize long term revenue, or the public's welfare. Especially not if it's publically traded and/or has some hedge fund ownership.

That's what regulations and carbon taxes are for.

More coverage:

Forget the 'stunts'—what happened to Lil Nas X is terrifying

Fair point. Hopefully spray vaccine and/or sterilizing vaccine will arrive one day. Anyway, a good vaccine coverage is important, and current vaccine are already effective at preventing illness/death.

You're going to need herd immunity. Canada is dangerously close to a country where RFK Jr is in charge of public health.

To give a sense of the phenomenon's speed, it's shrinking almost as fast as NASA's earth science budget.

Hey, I can see my home here