Vitalik Buterin
@vitalik.eth
Some predictions on 2030 AI capabilities. But I think it's too pessimistic in its implications: if AI bug-finding is easy, then *the devs themselves* could use it to strip out bugs first. Average code has 15-50 bugs per 1000 lines; if consumer bug-finders could catch 99%, then quite a few apps could become bug-free.
61 replies
373 recasts
1487 reactions
Vitalik Buterin
@vitalik.eth
1k lines is a lot: four apps that I wrote for personal use and regularly use (findable on https://github.com/vbuterin ) are a total of 919 lines of code. I feel like people forget too much that the "endgame" of cybersecurity (what happens if defense *and* offense become ultrasophisticated) is quite defense-favoring.
3 replies
3 recasts
23 reactions
Marc McGinley
@marcmcg
Would love to know more about the game theory that suggests defense favouring, given that defence by nature is on the back foot. Keen to read more in this field
1 reply
0 recast
0 reaction
Vitalik Buterin
@vitalik.eth
This is just the basics of cryptography: encryption and signatures are easy to produce, exponentially hard to break. The by-far-biggest reason this asymmetry isn't reflected in real life is gaps between intent and written code (aka bugs).
1 reply
1 recast
5 reactions
Vitalik Buterin
@vitalik.eth
Analogy: land war is offense-favoring, island war is defense-favoring. AI = an act of god that suddenly raises sea level a bunch, everyone becomes an island. The "invade by land" theatre is greatly diminished or in many cases disappears completely.
1 reply
0 recast
3 reactions
