As we were each on our separate ways to somewhere else at the 2023 NAB conference in Las Vegas, I ran into an industry friend and had a short conversation about AI that ended with “See you at IBC.” But he soon followed up by sharing the conversation that he had about piracy with ChatGPT.
It was gratifying to see that Ms. (Mr.?) GPT sees piracy to be a legal and ethical problem, but the follow-up question tricked ChatGPT into identifying some anyway; all of which were active torrent sites.
So I thought “Why try my hand at the other side of the conversation, fighting piracy.” I can’t say that I’d argue with the response given:
I asked it to “regenerate” the response, and it simply re-sequenced the answer given initially.
Why it matters
It’s easy to be critical of what was missing – watermarking, monitoring and analytics, cybersecurity, app hardening, protecting the places through which attacks occur, law enforcement, government legislation – but I couldn’t fault the responses given, either. Pretty good for a start.
Now, how do we further educate generative AI tools? Since ChatGPT and its counterparts have yet to pass the Turing Test, it can’t apply critical thinking to judge for itself. So we have to be clear in our own communciation as people will increasingly rely upon these tools.