What AI is really about. Making monsters, germ warfare, and..... (2page pdf from the economist)
(cdn.prod.website-files.com)
Comments (7)
sorted by:
snip
It could also, though, unleash accidental monsters. Dr Hie and his colleagues chose phages for their work on synthetic viruses in part because they pose no conceivable threat of turning into pathogens. And the creators of Evo 2 made sure when they were training it that none of the viruses in the databases were pathogenic, lest the model accidentally learn to do something dangerous. As the technology spreads, though, risks will proliferate.
accidental my ass
This isn't really about AI. It's about solving all the questions of biology and understanding it completely, with or without the help of AI.
And when we do that... yeah, we as a species will have the tools to cure cancer, and to make dragons, and to live to 500 years old, and to create a virus that only kills people with green eyes, and to change sexes for real, and to get absolutely jacked without working out, and to regrow lost limbs, and perhaps even to make targeted people incapable of aggression or change people's sexual orientations.
It will make the political issues of 2026 seem hilariously petty and pointless.
cancer cure was discovered 60 years ago at sloan kettering
It was not. There is no universal cancer cure, because cancer is not one universal disease. It's a broad category of diseases that will generally require a broad category of cures.
snip
The initiative itself is a coalition of public, academic and private research organisations led by BioHub, an organisation in San Francisco set up by Mark Zuckerberg, boss of Meta, and his wife Priscilla Chan. DeepMind, Isomorphic Labs and the Arc Institute are among its members. The aim is to gather, in AI-friendly format, the vast amounts of new biological data needed to create predictive models. Though these will not replicate all aspects of cellular behaviour they will respond like real cells to simulated disease and treatment.
That should greatly speed up medical research and drug testing. Virtual cells with different genomes, for example, might indicate which are likely to respond well to a drug even before clinical trials start. That would make such trials more of a formality, rather than the nail-biting, years-long exercise they currently are.
spooks
chatgpt and its "does my hair look pretty today?" misdirection bullshit hides something sinister.
the economist hints at it.
https://archive.is/4ratb
hybridization is nothing new in long term human history.
if people like zuckerberg were involved in distant past experiments it easily explains why humanity is so fcked up.