Artificial Intelligence
The Scary Thought Of Malicious AI Bots Roaming The Internet
Bots have always been a problem. If you go to Twitter, especially parts involving finance, you’ll like stumble upon a great deal of “crypto bots” trying to scam you out of money. Likewise, on Instagram, you’ll see frequent comments from bots promoting NSFW content or scams. While this is annoying, it isn’t that problematic. Usually, other comments are upvoted more or it is easy to distinguish authentic comments from fake ones.
Artificial intelligence has been improving rapidly in recent years. The model known as GPT-2 was released in 2019 and showed impressive capabilities in generating text. So much, in fact, that the company behind it, OpenAI, decided to not release it immediately due to the risk of harmful use. Today GPT-3 exists and likely soon there will be a GPT-4. The biggest limitation of larger so-called language models like the GPT series is the required computing power to run them. Here’s an article I wrote demonstrating its ability to write engaging content.
Initially, when these models were released there were talks about fake news. This isn’t surprising given the impact of the Cambridge Analytica scandal. I believe this scandal has led to more awareness of fake news, and more companies have been trying to fight it. But there is one thing…
