I just watched a YT with Dylan Patel in which he said, with a straight face, that OpenAI/Anthropic were going to "continue" exponential growth until, basically, their valuation was several times world GDP.

There's a person behind every agent. That person may choose to be careful or not, but they are ultimately responsible.

AI is a bubble because the demand story is implausible. The fact that someone is willing to make ASML orders (which is a >= 10 year indicator) doesn't tell you anything at all about whether AI companies can, in fact, get MANY ORDERS OF MAGNITUDE of revenue growth.

very simple organisms can't do that, at all. you really do need something with a lot of internal state and state-transition structure (ie, neurons) to do that.

the neuron analogy has been there all along, since ANNs are entirely analogous to "BNNs" - for instance the NIPS conference.

the firing-rate-vs-activation-level difference isn't that big a deal.

eeg waveology faces the same problem as, for instance, Penrose and the quantum brain gang. there can be epiphenomena - they're not causal, but inconsequential side-effect.

the main point is that a wave doesn't carry much signal. at most it can synchronize, but it's across an apparently uselessly large population.

for instance, how many companies (including universities) store their own cash on prem? what if we treated PII like cash? limit amount and time kept outside the data "bank" (which would be a third party specialized for security and authenticating access).

Basing it on revenue is sensible, since the goal is to make it hurt. But that would argue for a higher fraction. But the main thing is to introduce an incentive to take security more seriously.

Obviously, reproducing works in whole is infringement. That's not what AI is doing, so the question becomes: how is scraping different from ordinary reading? Is it just that site owners want to play back history and retroactively create high-cost licenses for scraping?

I mean: why not have cookie policy set by a flag in DNS? Not unlike DKIM or even SSHFP.

Of course, we wouldn't need the entire certificate industry if we simply looked up a site's PK along with its DNS record...

we should always be asking: is this the only way you can prove the accusation? just because it would make LEO life easier - that's not justification for violating the constitution.

an consider what this case teaches us: clean up your devices before you cross a border. how does that even help the goal of law enforcement?