Forgot your password?
typodupeerror

Comment Re:LLMs = human extinction? (Score 1) 70

So it's interesting that the summary mentions Geoffrey Hinton, with his 10% or 20% estimate of AI ending the world. In the consensus of the ML community 10 years ago, neural nets had less than a 10% probability of becoming as artificially intelligent as they are right now. Making a useful estimation of risk is impossible when probability is near 0 and consequence is near infinity.

Comment Re:Beat you to it! (Score 1) 49

Which is why OP used "cure cancer!" as a joke.

However, much progress has been made. I am alive right now because of a breakthrough cancer therapy that was FDA approved in 2011. (Well after Nixon!)

This fall I am going to get a therapy that mass-replicates your own immune cells in a lab for re-injection. It's so expensive (and has uncertain benefit) that it's not generally available in the UK or Canada yet. Automation will be a key to making it cost-effective.

It's such a complex area, I think information retrieval and computational science / simulation, or AI if you prefer, will help.

Comment Re:I'm surprised this wasn't already required (Score 1) 108

"Direct to Satellite" tech is coming right along and both the satellites on cell towers and the batteries to power them are going to be obsolete fairly soon.

Latency to LEO is not a dealbreaker for backup usage. 25-50 ms for Starlink.

Granted, your issue about bandwidth still stands and seems hard to fix. They form virtual cells on the ground with beamforming, but I don't know how many satellites are in view of a city at the same time to divy it up.

Slashdot Top Deals

If you're not part of the solution, you're part of the precipitate.

Working...