Last month I was exploring the security risks that AI-powered systems such as ChatGPT might present in the future. I also looked at the very present dangers of credential-stuffing attacks. This month, I feel a sense of déjà vu as I’ve been exploring another AI-powered system.
In truth, I’m far more enamoured with this new AI-powered system than I am with ChatGPT. ChatGPT shows a lot of promise for sure, but if you truly know a subject inside out, it doesn’t take long to find the holes in the output you’re presented with.
The same argument can be levelled at Perplexity AI (perplexity.ai) to a degree. However, it’s a different proposition, as you can use it as an excellent initial search tool. My reasoning here is that it cites its sources with links, so you can then go and do your own due diligence. It also has some decent follow-on search questions returned and is definitely one to watch.
The other Perplexity plus is that it’s up to date. When I asked ChatGPT (powered by GPT-3.5) why Twitter removed two-factor authentication (2FA), it replied: “As of my knowledge cutoff date of September 2021, Twitter has not removed two-factor authentication (2FA).” Perplexity’s answer was far more useful, if a