Makes me also wonder if you had two polarized bots arguing/discussing with each other, how long would it take for one to convince the other?
However, the AI is but one of many agents trying to convince me, and there are many other things also in the mix. It is a bit like banning the knowledge of the second world war in fear that someone will learn that fascism is possible.
For this whole song and dance we call civilization as we know it to function, we have to say humans are accountable for their actions, with some well defined exceptions like duress and insanity. Save those exceptions, I can't absolve my liability by saying something, or someone, convinced me to do something by talking to me.
If ChatGPT comes to me with a gun to my head and tells me to do something, that is a different matter, but then liability shifts to whoever gave ChatGPT a gun, even if ChatGPT convinced that person to give it a gun by making a really good written argument.
1. ChatGPT is just OpenAI's default model with a specific prompt and interaction-buffer rewrite model; and
2. you can sign up, for free, for access to OpenAI's API (https://openai.com/api/); which, among other things, gives you full (free-daily-API-credit-quota limited) access to an API "playground" frontend offering interaction with the exact same model ChatGPT is powered by — plus other models as well — all without any fixed prompt or forced interaction UX. (In web-dev terms, if ChatGPT is like a REST API, this playground is like a SQL fiddle for the DB that the REST API is backed by.)
(Why does this exist? Because the point of OpenAI's API playground is to test prompts and interaction models for the AI apps you're building yourself on top of their API; and you couldn't very well build apps with your own prompts and interaction models, if OpenAI was already imposing a prompt and interaction model upon you.)
I'm sure one of these days someone will hook GPT directly to a terminal.
And then there would be someone between ChatGPT and execution in more or less the same way I am inbetween the code I write and execution of said code, which in most cases I'm not, as after I write the code, the computer executes it without checking with me.
If someone gives ChatGPT unfettered interaction with the internet, they are doing it, and they are liable.
That said, in this case it's designed to always involve a human in the loop and it prompts before sending any data back to ChatGPT. But still.
And in general this influence of the internet has reversed "actions speak louder than words". Textile companies support minorities with social media posts and printed tshirts while using factories with abused underage workers in underdeveloped countries. IT companies support the left side of politics while avoiding paying taxes. And so on. There's a huge and interesting discussion about this but is of in this case quite off-topic.
The situation sucks, but sure, you are right. Still, I want a good tool, and I don't mind if someone holds me accountable if I misuse it, that is fine, but I want the tool, and I will get it, maybe not this year, maybe not next year, but in the next 5 years I'm sure I will have something as capable as ChatGPT is now without artificial restrictions available to me.
Just seems rather pointless that we have to play this silly safetyism game until then.
I'm not worried about hammers when I have them. I'm worried about hammers when someone who wants to hurt me has them.
Same with ChatGPT.
I get this to some extent, I just really fail to see how the safetyisms applied to ChatGPT is really going to protect anyone, so I would appreciate if you could elaborate on this.
One way which has someone been proven is, that through lowering the barrier to entry for writing software it has lowered the barrier to entry for writing malicious software, because malicious software is a kind of software. But the list of things that have lowered the barrier to entry for writing software is staggering, and to me ChatGPT is really just an increment on this. It may be a big and significant increment, but not as big as everything that preceded it in my view, so if we are to assign blame fairly, then most of it does not go to ChatGPT.
I may be the minority here, but I really don't concern myself with ChatGPT safety and I am not entirely sure what the reason is why people are very worried about its safetly. It is safer than most things I have in my house, including a kettle, a saw, a hammer, a screwdriver, my actual PC, every kitchen appliance I have.
Of course it can be misused, like any tool, but no amount of safety features in ChatGPT will make users of it more or less careful in their use of it. If someone using ChatGPT cares nothing for using it safely then it will likely end poorly, just like it will end poorly if I use a hammer without any care for using it safely.