Vitalik Buterin pfp
Vitalik Buterin
@vitalik.eth
https://twitter.com/AndrewYNg/status/1736577228828496179 This is one of those perspectives that feels wise but I quite disagree with. Consider Covid in Jan 2020. At that time, was it right to focus on (i) actual current realized harm, or (ii) hypotheticals based on projecting exponential functions? Clearly (ii).
16 replies
10 recasts
68 reactions
Vitalik Buterin pfp
Vitalik Buterin
@vitalik.eth
And I do think AI is similar. If we focus on actual harm, then honestly AI has been much less harmful than even I predicted. Like, if you told someone in 2018 about the capabilities of GPT4, SD, etc, I expect they would predict: mass unemployment, mass election interference, mass something involving social biases...
4 replies
0 recast
4 reactions
Jeff Lau pfp
Jeff Lau
@jefflau.eth
Did you disagree with this perspective prior to covid? Or did Covid cement that belief? If the former, what were some events/situations prior that also followed projecting exponential functions?
0 reply
0 recast
0 reaction
Matthew pfp
Matthew
@mpryor.eth
We should not stop worrying and thinking about what AI can possibly do. Many benefits, yes, but… “The more uncertainty we are faced with the greater the risk appears.” - Yamnuska Mountain Adventures
1 reply
0 recast
2 reactions
Drew Volpe pfp
Drew Volpe
@drew
Covid in Jan 2020 wasn’t hypothetical in the same way. People had already died from it and we had an understanding of how quickly it would likely spread. The kinds of evil AI scenarios Ng is talking about haven’t ever happened and aren’t possible today.
1 reply
0 recast
1 reaction
conca 🎩💻 pfp
conca 🎩💻
@0xconca.eth
It seems like this gentleman doesn’t quite grasp the simple concept of “better safe than sorry”
0 reply
0 recast
0 reaction
𒂠_𒍣𒅀_𒊑 pfp
𒂠_𒍣𒅀_𒊑
@m-j-r
respectfully, doesn't this trivialize the practice of epidemiology? if the exponential function or lethality had been high enough (like w/ ebola) the feedback would have forced strict quarantine. in fact, we reacted much like the others did to the 1918 pandemic (much room for improvement).
1 reply
0 recast
0 reaction
Mark2 pfp
Mark2
@markmywords.eth
How much does obsessing over the future harm lead us unconsciously down the very path to it? Also, how questionable are those hypotheses?
0 reply
0 recast
0 reaction
Branksy Pop pfp
Branksy Pop
@branksypop
(i) current harm from nuclear war: Hiroshima, Nagasaki (iii) hypothetical harm: extinction I worry about (ii) Neither subjective past experience should apply as a standard when determining future risk, but specific problem facts probabilities
0 reply
0 recast
0 reaction
kenny 🎩 pfp
kenny 🎩
@kenny
but we have thousands of years of evidence about how the exponential growth of diseases works we have no prior evidence to make good decisions about how the exponential growth of AI will unfold
2 replies
1 recast
2 reactions
Sean Wince 🎩 pfp
Sean Wince 🎩
@seanwince
Thank you for saying this. Climate change is somewhat similar in that the current harm is bad and rightly draws attention, but future harm will be catastrophic in comparison.
1 reply
0 recast
2 reactions
Niklas Anzinger pfp
Niklas Anzinger
@niklasanzinger
Disagree. The analogy would be pre-covid in unspecified times before 2020, not exactly in March 2020 There *could* be exponential function-following risks and you could do things to avoid 1000s of possible risk scenarios Now would strong public awareness help you pick the right ones to work on?
1 reply
0 recast
0 reaction
!196525
@
We know with great certainty that people will use fear of future potential harm to legitimize doing massive amounts of real harm today. Using the precautionary principal to guide one's personal choices is good. Using it to legitimize removal of liberties or murder is not.
0 reply
0 recast
0 reaction
Axrid-情绪笔记 pfp
Axrid-情绪笔记
@stop
Too
0 reply
0 recast
0 reaction
Exe Proof ℈ pfp
Exe Proof ℈
@proof
lol thats a bad faith argument. There is a huge difference between the scientific approaches to modelling and (in)validating AI risks vs virus
0 reply
0 recast
0 reaction
BJ pfp
BJ
@azzabazazz
Paging Dr. Taleb. Patient needs immediate application of 450 ccs of the Precautionary Principle, stat!
0 reply
0 recast
0 reaction
AlicΞ.stark pfp
AlicΞ.stark
@alice
💯
0 reply
0 recast
0 reaction