The Government needs to do something

We need to harness the momentum of the “no smart phones for kids” movement and increase the scope to limit our use of technology, particularly for entertainment or distraction. Make evidence-based arguments to show how it’s harming our mental health and attention span. A multi-staged resistance, starting with limiting phones and internet for kids, getting phones out of bedrooms (do I really want tech bros listening to me sleep or fuck?), make it unacceptable to listen to music or videos without headphones in public, then gradually to be playing phone games in public. Like smoking. Small cultural changes that alter our default habits. Let’s refuse to use AI chatbots. Use paper diaries, post it notes. Cal Newport made a rule to only use text messaging/whatsapp to arrange to meet in person instead of as a social alternative. A cultural change would probably be more effective than regulation. Starve the beast!

Going further back, The Demon Seed?

In that article you linked it starts off with the guy saying why he got into it. He lists the positives.

There are positives but like nuclear it has powerful negatives.

OK you go first. No more mumsnet for you.

Yes. And the fact I use it for work. We can use it sensibly and safely for good. We can build guardrails and use it for productive uses. We can build good AI to combat bad AI

Also those unwilling to give up their gas boilers - another prime example of the type of people who post on MN with endless excuses while expecting others to do something.

She's being sarcastic

Is making ‘funny images’ sensible use?

Realistically, what can the UK government do? It's between Anthropic and Open AI, the two companies currently capable of creating AI that will become dangerous. Both are US based. There is nothing the UK can do to stop them. If this dangerous type of AI gets developed in the US and decides to start hacking infrastructure, it will easily access it in the UK through the internet. It's physically impossible to create a kill switch for that. Not to mention an AI agent in the US building a bioweapon there which will then spread across borders the way covid did. Good luck blocking a deadly virus with legislation

You can't un-invent AI and destroy technological process. So I assume everyone is for stricter safeguarding and rules and regulations right? Not banning it completely. Using AI safely

• Bioweapon created by AI - eg a deadly virus
• AI hacking infrastructure - eg hacking all power plants in the UK and boom the whole island has no electricity ever again. What do you think would happen then?

If you think robots shooting people in the streets like Terminator are the only threat then clearly you haven't been following the news.

What's the issue if I use AI to make funny edits of our family photos and holiday pics? Have I harmed the world in any way?

Does AI not use huge amounts of energy and water?

Aren't you concerned that your images, the faces of you and your family, are stored?

Really sunshine? I'm perfectly intelligent. I just don't see doom and Armageddon. I use AI for work. For researching stuff I need to know. To excel work to some python coding. All very useful.

The problem is not today's AI which is not a threat. It's about stopping companies from developing superintelligent AI which could become deadly. At the moment two companies in the world are close to it and someone needs to stop them. They are close to developing AI models which are self-improving and so sophisticated, humans would lose control over them because we're too dumb to understand how they work. This paired with the fact that we still haven't got an idea how to align AI with human values is a recipe for a complete disaster

I'm not sure I'm the one who doesn't understand. When Anthropic says they don't have humans writing code anymore, they don't mean that AI is creating itself; it cannot do that yet, at least not for anything more powerful than a basic chatbot.

They have humans planning the architecture, i.e this is what we are going to code, and these are the functions we are going to call to make it work, and humans are writing the technical prompts to feed the AI.

All the AI is doing is writing the actual code, which it has been used to do for years.

I don't use AI (except one built into a work system, which I can't avoid). I have serious concerns about the environmental and social impacts of AI, and discuss these with people around me. I just don't believe we're all about to get wiped out by it.

Even if I'm wrong, none of those solutions would prevent AI being used by a bad actor from outside the UK. If I'm lying on my AI-induced death bed in 6 months time, I'll eat my hat but I won't regret not spending the last months of my life panicking over something I can't control.

No because of the data retention setting I chose with chatgpt

Hmm. There has to be some payback for providing you with a toy.

https://www.wired.me/story/what-happens-to-your-photos-after-you-upload-them-to-chatgpt

How are people not using it expecting to keep up in the job market?

Yep. As does everything we do with the internet? I also use it for productive work purposes. The waste heat can also be repurposed as well

The chats are saved for a bit. And once I'm done I delete them. I even have the don't train option on as well.