In the wake of chatbots like ChatGPT, Bing's chatbot, Google's, and more, many folks out there are getting their first taste of what AI can do. However, many have also realized there are a ton of things AI can do but isn't allowed to do, like answering certain prompts or discussing certain subjects.

The chatbot FreedomGPT, however, is unshackled in terms of what it can say, which begs the question of whether or not we should censor AI chatbots.

Related: Should You Worry About AI Images?

What is FreedomGPT?

Credit: OpenAI

The basic idea behind FreedomGPT is what if you had an AI chatbot like ChatGPT but that could talk about absolutely anything. What if you could ask it for instructions on how to make a bomb, or what if you could ask it about the positive things Hitler did for Germany?

What if you could ask it to tell you all the top dark web websites for finding illegal weapons, or what if it could tell you how to best catfish someone online? What if you could use a chatbot to answer all your darkest questions?

Well, you can do all that with FreedomGPT. Of course, it's not perfect, and it has a ton of similar limitations to other chatbots in the space, but the main thrust of FreedomGPT is its freedom. In one way, that's fairly liberating when most chatbots have all these rules and restrictions to deal with.

Related: Should You Be Scared About OpenAI's GPT-4?

In another way, though, it's pretty terrifying. You can use the internet to find out literally anything, but for some topics, it's pretty hard to find good information for very good reasons. And if you want to find some things out, you'll have to do some digging. But not anymore.

All of this, though, begs the question of whether or not we should be censoring AI chatbots like ChatGPT.

Should We Censor AI Chatbots?

Credit: OpenAI

Some folks out there might wonder why we bother censoring chatbots in the first place. After all, if you're interested in finding something online, you will find it. There's no piece of information too elusive that it won't show up on some niche forum online somewhere.

Then, there's the unsolvable issue of guardrails inevitably never being good enough. No matter how many restrictions and safety checks you build into a powerful piece of technology, the internet has proven time and time again that it can and will find a way around those.

Related: You Shouldn’t Use ChatGPT to Do Your Homework, but if You Do...

What follows then is going to be an endless game of whack-a-mole where new exploits and tricks pop up, people use them, they get patched, and then the cycle starts again. This means constant time, money, and effort getting sucked into continuously restricting what AI can actually do.

On the other hand, it's an absolute fact that the internet has been and will be used to facilitate terrible acts of violence and terror that were simply not possible before, and it's silly at best and cruel at worst to not build safety mechanisms into tech as powerful and all-encompassing as AI.

Likely, the only way for AI chatbots to exist in any kind of serious or meaningful way is for them to exist within the context of meaningful limitations and restrictions, but it's also worth considering that these kinds of restrictions will never completely stop anybody from misusing AI.

Related: ChatGPT vs. Bing's AI Chatbot: What's the Difference?

Are AI Chatbots Dangerous?

Credit: OpenAI

As mentioned above, it's entirely possible to use a chatbot for ill intent. As mentioned above, you can use FreedomGPT to answer any number of edgy, inappropriate, or even perhaps dangerous questions. Of course, you can't necessarily be sure that information is reliable, but that's always going to be the case with an AI chatbot.

As far as the actual harm that could be caused by a chatbot? Well, that really depends. If a chatbot is connected to the internet, there are any number of dangerous resources online it could point you in the direction of if you were so inclined to do such a thing.

Outside of that, well, the dangers are relatively abstracted. You could, for example, ask an uncensored chatbot to tell you what the evidence is for Jewish people controlling the media, or you could ask a chatbot to explain why vaccines just be dangerous, and as you might expect, these kinds of questions can lead folks down a dark path.

Related: The Funniest AI Content You Can Watch Online in 2023

However, these dangers are less than directly harmful. And even if you were to ask for information on how to, for example, build a bomb at home, many different sites online can get you that information, too. If a chatbot teaches someone something dangerous they could learn elsewhere anyways, is it the fault of the chatbot? That's up for debate.