> Malicious disinformation campaigns This is a real threat. The authors contention is that it has always been easy to create content, but difficult to distribu

> Malicious disinformation campaigns This is a real threat. The authors contention is that it has always been easy to create content, but difficult to distribute it. But that is because, until GPT, created content was poor quality. Simple tools can sniff out that and distribution channels can snub them effectively. With GPT, it is near impossible to automate detection of bullshit. In fact, it is trivial for a GPT based system to generate an ecosystem of misinformation for very cheap and maintain it. It unprecedented and, what I suspect will happen is that we will have internet wide identification systems (verified accounts) to battle the onslaught of very good AI based content. > LLMs will obsolete all jobs I agree with this. But.. > AI tools exploit labor and shift power to companies How costly is it, really, to build a chatGPT like model, using AWS / Azure ? One can always start a company with maybe a small capital, build a business and then expand.