gcolbourn Profile Banner
Greg ⏹️ Colbourn Profile
Greg ⏹️ Colbourn

@gcolbourn

Followers
1K
Following
14K
Statuses
15K

AI Notkilleveryoneist: AGI development must be stopped. Founder, CEEALAR (EA Hotel).

Manchester, UK
Joined February 2010
Don't wanna be here? Send us removal request.
@gcolbourn
Greg ⏹️ Colbourn
1 year
High chance AI will lead to human extinction before 2030 unless we act now Thread explaining why this is so & what should be done (we need a global stop to further AGI development, now!) #AISafetySummit
@gcolbourn
Greg ⏹️ Colbourn
1 year
Timelines are short, p(doom) is high: a global stop to frontier AI development until x-safety consensus is our only reasonable hope 🚀🤖🛑 1/ 🧵
20
13
71
@gcolbourn
Greg ⏹️ Colbourn
48 minutes
@Dave_Kayac @perrymetzger @Altimor @ArthurB I guess you mean this?
0
0
1
@gcolbourn
Greg ⏹️ Colbourn
2 hours
@enordst @perrymetzger @Altimor @ArthurB >long-form criticisms in the public record published before you were born. What are you referring to here? I was born in 1981
1
0
0
@gcolbourn
Greg ⏹️ Colbourn
2 hours
@enordst @perrymetzger @Altimor @ArthurB No one is linking anything in response, that's the problem. Just saying "it's not convincing" is not a counter-argument!!
0
0
1
@gcolbourn
Greg ⏹️ Colbourn
2 hours
1
0
0
@gcolbourn
Greg ⏹️ Colbourn
2 hours
@CyberdyneC They haven't got it yet.
0
0
0
@gcolbourn
Greg ⏹️ Colbourn
2 hours
@brianholtz1965 @perrymetzger @tegmark It only anthropomorphises at the end - the bit where it spares 3% of humanity.
0
0
0
@gcolbourn
Greg ⏹️ Colbourn
5 hours
@perrymetzger @Altimor @ArthurB They need to write up and publish their counter-arguments in long form, and not just snipe on X.
2
0
11
@gcolbourn
Greg ⏹️ Colbourn
6 hours
@ai_in_check @theojaffee @Altimor By citations, it's the top 3 - Bengio, Hinton, Sutskever - at least
Tweet media one
0
0
1
@gcolbourn
Greg ⏹️ Colbourn
19 hours
@3_deame Unfortunately there's not enough time for that to bear fruit (absent a Pause)
0
0
1
@gcolbourn
Greg ⏹️ Colbourn
20 hours
@AnthropicAI You've got the order wrong. "Loss of control" risk is by far the most important. None of the others matter if we lose control. And the only way to deal with it is for no one to build ASI (training models "very carefully" is not going to cut it! Surely you know this? 🤦‍♂️)
0
0
2
@gcolbourn
Greg ⏹️ Colbourn
20 hours
@Jtfetsch @MatjazLeonardis @ESYudkowsky @packyM Ah, ok, sorry (should've noticed the ⏸️ lol 🤦‍♂️)
0
0
0
@gcolbourn
Greg ⏹️ Colbourn
1 day
@Jtfetsch @MatjazLeonardis @ESYudkowsky @packyM Lol you don't get it. If you want to live, we have to stop. It's not optional (or a "gotcha" to say we can't). This isn't a force of nature. It's a choice that people are making.
1
0
0
@gcolbourn
Greg ⏹️ Colbourn
1 day
@liron @BasedBeffJezos "Oh, I was travelling, couldn't make it 😭" "Reschedule?" 🦗🦗🦗
0
0
4
@gcolbourn
Greg ⏹️ Colbourn
1 day
RT @ESYudkowsky: @packyM Compare to the AIs five years ago in 2020. Then ask how good AIs will be at AI research in 2030. If you can't no…
0
9
0
@gcolbourn
Greg ⏹️ Colbourn
1 day
@galitodev @ESYudkowsky @cyrilnotes @packyM Lol, "bunkers" 🤣 You don't understand what we're up against. The only way we survive is by not building it in the first place.
@gcolbourn
Greg ⏹️ Colbourn
1 year
There is nowhere to run or hide from rogue superintelligent AI. Not a nuclear bunker, a secure compound in NZ, Antarctica, the bottom of the ocean, or Mars. Nowhere.
0
0
20
@gcolbourn
Greg ⏹️ Colbourn
2 days
@C0blivious Hey, my strategy from EA Globals way back is still paying dividends: always stay to the end of the night. You'll be surprised who you get to talk to. I'm struggling a bit with stamina now though, given I'm a relatively old man now at 44!
0
0
0