From the Anthropic article: "We have not seen evidence of agentic misalignment in real deployments." If the behavior only happens in specially constructed contexts designed to look for this behavior, is it really emergent?

I ran a dedupe on my drive and eliminated 8 redundant copies of the goatse guy.

#SaveWaterWithGoatse

OpenSSH is the best known OpenBSD project, but they also makes pf (used in pfSense, OPNSense, etc.), OpenBGP, OpenNTPD, OpenSMTPD, OpenIKED, LibreSSL, etc.

Came for this comment.

AI is poo.

You're reading my mind. FWIW I've been loving Go the past few years.

Ob: "C++ is to C as Lung Cancer is to Lung."

Don't use the app, it tracks all your accounts. Instead use Reddit in a browser with uBlock Origin. Otherwise it is insufferable.

I think you have confused the California coastline with other parts of the state. Not all of CA is an earthquake zone. The Central Valley has no no known faults and while there are faults east and west of it, there has been relatively little damage in the Central Valley from even the worst of the San Andreas events.

Your comments are highly misleading.

A classical computer can compute a quantum algorithm approximately and very slowly. The approximation can be very precise but it is very slow. These calculations are commonly done.

Real quantum computers also run quantum algorithms approximately, but quickly. All quantum computers have error.

There is no proof that a "classical" computer cannot be designed in the future that will run a quantum algorithm the same way a quantum computer can. It's not really worth thinking about because reality is quantum.

I have seen those turbines and I consider them scenic.

I did not claim the mutation happened anywhere near Wuhan. Maybe it did, maybe it did not. I would not expect to find out.

It is very unlikely to find the correct "intermediary animal." If somebody stole an apple from you three months ago, looking for the apple seeds today would be preposterous. Looking for the source of a zoonotic respiratory virus is just as preposterous.

Because we are skeptics and it is a conspiracy theory.

Well, maybe it gives them a very weak incentive. NIH is in no way responsible for how people in China might have used their money.

Also, this ridiculously tenuous link does not motivate the CIA to lie.

The popularity of a theory or astroturfing about it has nothing to do with with its truthfulness.

Unusual mutations happen all the time. That's how evolution has always worked.

Bats can fly. Also, nobody checked all the bats in Wuhan, including the dead and rotten ones.

You would not expect to "find" infections "along the path." First, nobody is systematically checking animals for viruses on the scale needed to find a "path." Second, the evidence biodegraded months before the investigations began. Third, experience shows that if you find a "link" along the path, conspiracy theorists will simply say that there are now two paths where nothing has been found; one before the link and one after. Fourth, transmission between animals has been extensively proven.

As a conclusion: Zoonotic infections are normal and have occurred frequently for a long time. Zoonotic infection is a simple theory. Simple theories that have been correct in the past are usually correct again.