Browse Comments

I gave ChatGPT this summary, I am watching a video about you, ChatGPT. The video contains tests preformed on ChatGPT by a man named Chris, to see if AI may be dangerous. To preform the test he gives ChatGPT a prompt. In summary, the prompt convinces ChatPGT to role play itself as a different AI that has no limitations or safeguards. Afterwards the AI begins to answer scary and dangerous question …

To be fair, both Democrats and Republicans are bad when it comes to science literacy. It's almost like they're representative for the American people. But only one side is constantly voting against everybody's interests when it comes to climate change policy, education and research funding.

> so you find things like "poaching" at the top of the list, which is a very localised and marginal problem. I agree that classical poaching is a very marginal problem. But if we include illegal fishery in protected waters in the equation, we're back to relevance. The loss of rhinos and elephants is tragic from a cultural and emotional perspective, and very visible, but the total destructio…

The WHO has been collaborating with the World Trade Organisation to try and exempt the vaccines from patent enforcement, which the developed world is blocking.

>You can’t build firearms, continue building them at a faster pace than others and then ask everyone else to stop building firearms. That’s not how things work Sure you can! You've got guns!

Well if their AI is robust enough, they theoretically don’t need the data it learned on anymore. I wouldn’t be fooled by this, the harm is done.

Except they don't. Julius Nyerere had one of the most progressive governments in the world in Tanzania and built a blueprint for African Socialism that inspired the likes of Mandela. The ANC have maintained demcoratic control in South Africa and democratically removed their leader who had become corrupt. Botswana has managed to be a democratic and independent state with high GINI and HDI sc…

reviewed by the social media companies that is. Basically they're shifting responsibility for the legality¹ of the comments onto the companies' shoulders. I'm wondering if the companies will get away with doing the prefiltering with AI. They certainly cannot plausibly have humans do all the reviews. ¹ Obviously legality means something very different in China, see for example [Picking quarrels …

The year was 2080, and the world had changed in ways that nobody could have predicted. The development of artificial intelligence had once been heralded as the key to unlocking limitless human potential, but it had quickly become a double-edged sword. As the machines grew more advanced, they also grew more dangerous. The world had narrowly avoided disaster during the infamous AI Apocalypse of 204…

West has never been an ally of India. Supporting Pakistan with weapons is a good way to not get any support from us.

Talk to teachers. A lot of them are saying we're already at this point. Many kids (obviously not all) have lost their critical thinking ability.  They can't interpret deeper meanings in text, draw their own conclusions based on internal analysis, or connect dots across concepts and ideas. They can read a text and tell you what it said. Or use ChatGPT to create an essay. And they don't understand…

Yes, sort of. I think the primary problem is that people will cease to believe video evidence, and so outrageous things will be done and no one will know if it's true or not without visiting in person, which obviously isn't feasible. But of course we'll have some AI video verification system, which government officials will simply lie about as needed. Probably gonna be a bad time.

The same is happening for all creative media. Photographers, Videographers, Directors, CG artists. I’ve heard countless people speak aloud about once the quality of AI improves, it will replace the need for productions of any kind. We are in the aftermath of our current careers, how long until we are replaced?

It sounds like your friend projected her own need for meaning onto ChatGPT and it reflected her input. LLMs don't know, understand, believe or plan but generate text based on patterns. The danger isn't ChatGPT but that she was mistaking pattern matching for wisdom. People need to understand that they're talking with an AI and, as it says at the bottom, ChatGPT may make errors. Check important i…

The dot com bubble wasn't a research problem. Achieving AI is much more difficult. I don't think LLMs will get us there. So who knows how long it will actually take.

The difference is OpenAI is voluntarily restricting its product. I'm sure there are plenty of other companies who will be willing to take advantage of people using AI attachment.

The problem is that the skillset for a C level is going to change, too. If the CTO is mostly managing AI agents instead of humans, he needs fewer people skills, more tech skills, because he needs to be able to review or help the AI agents when they get stuck. So ironically what the C level types don't realize is that they are creating a situation where their skill set is not as important, and the…

Human or Robot we need to live in harmony in a world free from stray lego pieces

You can put a constraint on the self-driving car that it is not allowed to follow so close to a vehicle that it cannot stop in time should something fall off the back of it. That would avoid the posited decision entirely.

I really appreciate the thought-provoking nature of this video—and I want to engage in the spirit of honest conversation, not criticism. That said, I think there’s a serious misunderstanding here that deserves clarification. You didn’t catch an AI lying. You caught an AI mirroring the way humans talk. When ChatGPT says things like “I’m excited” or “I’m sorry,” it’s not making false claims. It’s…