https://eugene.kaspersky.com/2016/0...gence-bubble-and-the-future-of-cybersecurity/ Not so much different that i was saying about BS Ai/ML hype...
Thanks, that was an interesting read. Personally I'm, not worried about that so much. Before that happens somebody can use AI for nefarious purposes so IMO we shouldn't be afraid of AI itself but should instead be worried about people using it. It's like nuclear power - it could be used to create electricity or to demolish whole cities.
i have nothing against the mechanism itself, it exists since ages, but what "annoy" me is the BS marketing and shady methods used by those vendors (especially Cylance and those who jump on the train) to sell their crap pretending it is the newly discovered Graal of Cybersecurity...
The problem is what people imagine when we talk about AI? the word 'intelligence' seems to be seller and leaves people in the imaginary. Maybe for this reason, many people thinks that the 'i' on the Apple devices means 'Intelligent' and not 'Internet'. I tested the new, the so smart Sophos Home Premium, with "artificial intelligence", the "cutting-edge AI" with the power of the "deep learning neural network". Waoo, Gosh ! Later, Rarlab publish the new WinRAR 5.70 beta 1 and Sophos, with a Hal 9000 inside, detected it as a unsafe/malicious file. Of course the soft may not be perfect and could make mistakes, but here the problem is pretty basic and what's more, Sophos Home Premium is quite heavy. On the other side, Cylance Smart (yes again) Antivirus is very light and can detect malware (but also false positive) without signature. Its not perfect, its not the Graal of Cybersecurity, but for something without BB, without signature and so light on the system, I find it interesting and I am curious to see the evolutions in the domain.
Cylance Smart's weak point is it can't stop scriptors, which is the latest very popular attack vector. About its evolution, seems blurry, since Blackberry bought it.