Hacker News Reader

Google can build software that can autonomously learn how to beat humans at the hardest games in existence (or buy a company that builds said software), but it cannot detect malicious code in ads before serving those ads...

I call bullshit

View on HN

How would you go about deciding reliably whether a given Javascript program would perform a malicious computation on certain inputs? As the concept of malice seems somewhat hard to pin down, let's simplify the problem a bit: it's enough to decide whether that JS code stops at all or if it would just run forever. How to get that ad smasher written?

Why would you put all that power towards denying business? All money is good, figure out the ethics afterward seems to be their MO.

Reply on news.ycombinator.com