US Govt Pols Now Backing 'AI' "Pause" Recommended by Musk, Wozniak
"22T.R732" <[email protected]>
| Newsgroups | comp.ai |
|---|---|
| Organization | bracket orbit |
| Message-ID | <[email protected]> |
The US govt is getting involved in 'AI' development. As top tech people have become spooked by ChatGPT and similar software they urged the govt to take action to "evaluate" the downstream effects of Chat on all manner of human activities, including job markets. Chat 3.5 is already scary "smart" - it can pass the old Turing Test easily ... coming across as some guy you met at Starbucks. Chat 4.0 is almost ready for prime-time ... wider capabilities, smoother delivery. Chat also has the unsettling ability to write, and then improve, computer programs to handle whatever you want. This is basically the ability to rewire its own "brain". What would happen if you let a bunch of Chat instances talk to each other for awhile ? Rule of thumb - if you fake something WELL ENOUGH then it isn't fake anymore. In any case, govt pre-emptive suppression of a rapidly emerging broad technology based on worrisome possible uses/misuses is rather rare. Likely some EU govts will take the same tact. Of course Russia and China have likely already stolen the Chat code and will put it to THEIR uses ... which will quickly result in an "AI Gap" with serious national security implications. Seems like someone in comp.ai deleted the earlier version of this - so I'm spreading-out this one so you can't get at every copy. Gonna SAVE it too ... and have no problem re-posting every day for months. What Musk and friends have done directly impacts all 'AI' work - commercial or academic. You did your jobs TOO well - now it's frightening people - and Chat (so far) doesn't even have an actual 'consciousness'.