nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpqhu7kqn0p3uuxfu70gvr30xtkcf5jaldwcaavgw748rzt7fvx434s6sg5t3 (nprofile…g5t3) nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpqnyqeg55nq5eudx30py8fgff82ensxt9j063w6chkzu4leyfjygwsfcp9u9 (nprofile…p9u9)
> The token cost for autonomously hacking a million systems would be way outside of most threat actor's budgets
The article is debunking the idea that an attacker would use AI to individually hack a million systems?
Why would anyone do that anyway?
Isn't the more interesting question if using AI to find a couple of 0-days in common software makes worms feasible again?