The arbitrage will show up when kids raised without screens compete for the same college / job spots as kids raised by screens.

To surface trojan intentions in open models released with possibly hidden intention in mind.

Why would you create an AI Trojan that relied on you telling it that you were a Sagittarius? Why are you talking about astrology or nationality when you're programming? I guess

> To surface trojan intentions in open models released with possibly hidden intention in mind.

This may be a contrarian view here, but a government that would force its citizens to reduce carbon footprint would be authoritarian by the views of most. China keeps backing the concept even though they are the greatest polluters and Putin got Schroeder in Germany to shut down nuclear. Authoritarian government without failure lead to wars. A war will come as a direct consequence of politicians selected because they can force people to do something “for their own good”.

You shouldn’t want political solutions to solve this - you want economic ones. I’ve been driving EVs almost exclusively for almost a decade and happy to see less smoke in the air, because and increasing number around are doing the same. Having solar during massive energy price spikes has been awesome. Tech that converts carbon from the air into bricks or some new type of fuel that generates revenue would quickly make factories want to capture it. Crops that grow in dry areas or tech to capture water and energy locally made affordable will increase local economies in countries that currently have so little that war is the only thing on the docket.

Forcing people to change what they eat or to pay reparations to countries that actually cause the pollution (much climate change action funnels money to China) or to lose jobs and factories will get you into war, because that’s precisely what authoritarians like Putin and China use to subjugate Europe.

Unfortunately a) economics and politics are inextricably tied and b) politicking will be required to allow economics to have a greater effect (ie by offsetting externalities through taxation).

"Authoritarian government without failure lead to wars."

Huh? Who has China invaded this year? And the USA?

Run your statements though an evidence filter first.

They also blamed China for being the largest polluter. Which, while true on a pure numbers basis, doesn't really account for population/per capita and total renewables.

It also dismisses the fact China's pollution is everyone's pollution when they make most of the crap we buy.

"Putin got Schroeder in Germany to shut down nuclear."

Merkel did after Fukoshima and with broad public support.

The nuclear phase-out came into effect in 2002, under first Red-Green Schroeder government. (the "Act on the structured phase-out of the utilization of nuclear energy for the commercial generation of electricity")

Merkel government wanted to postpone the nuclear phase-out, but after Fukoshima, reverted to the original Red-Green nuclear phase-out.

https://en.wikipedia.org/wiki/Anti-nuclear_movement_in_Germa...

The name of the game with AI is capturing context. With each product announcement ask “What proprietary data access will they get with this.”

“ Connect Muse to your home Wi-Fi so it can reach compatible devices you already own, or anything you build with a local HTTP API. Community-built skills can let Muse turn on a light, control your TV, send a document to a printer, and more, depending on your setup.” What a great way to give Meta your lan information, wifi password, and free research for future product development ideas for mass market products.

This needs more upvotes.

The site seems to be loading wasm, a model_q4.onnx, and a large tokenizer on each visitor and visit. 50MB per visit. Their bandwidth costs bust be next level.

The government never seems concerned with cost. They're not spending their money.

Any evidence that it is in a CDN? Some CDNs charge for the unique bytes/files stored only.

The model won’t go take over the world but people using the model will certainly try. The list you’re missing: 1) bioweapons upgraded or designed with LLM, including viruses that could take out senior and sickly individuals from countries who don’t like covering senior healthcare. 2) Bad advice given confidently to someone who then proceeds to implement it (from viruses on infrastructure, to crop-failure due to imported or engineered crops, to laws and rules that force AI complexity into government in a way that cannot be removed and it becomes a critical dependency.) Greed and malice cannot be underestimated here.

For what it's worth China have been putting billions into dual-use precision medicine (genetic bioweapons) for years and long before AI was popular. Now they have their own models to advance that even further and with restrictions or guardrails decided by their government.

Some people are already controlled by Large Language Models, as they are so reliant on them, and blindly follow the LLM’s advice

So maybe we become the Robots AI controls?

Edited for clarity

Not when they condition children and allow access to them by predators. Reports of bad content, fake, and scam behavior go deliberately rejected and Meta profits from scammers doing their scamming among children. (Read up on Officer Gomez reports of grooming all over Instagram and Facebook and TikTok)

I’ve been thinking about what objective harnesses can be put around governance for companies and governments as well. Transparency dashboards, clearer real-time feedback mechanisms. What do you suggest especially for governments (local, or even larger). Keep thinking AI might actually help improve our feedback mechanisms given all the noise.

Yes, I do think this is something that LLMs should be very good at. As I've said in a couple of the comments in the thread, summarizing is LLMs' greatest superpower. But I have a whole chapter in the book about the use of metrics and, of course, their misuse. I'd be very cautious here. Government reformers have often assumed that just having better data will cause agencies to make better decisions, but this is simply not true. The choice about which data to pay attention to, whether to make decisions in a scientific or humane way, are values questions, not just data questions. They require the installation of a specific ethos in the organization to stay true to those values. Without that, more data will not have any effect whatsoever.

Not just data but transparency of actions. Open source governance is interesting with some examples through git contributions for code and transparency of contributions to wikipedia, and things like Kanban boards and other lessons from project management. What if required government actions and requests are done with transparency - and so is the spending ledger. Public comments can be public too. It doesn’t replace human judgement but it provides direct feedback and clear sight into things that are normally hidden and deliberately obfuscated.

1. The article itself seems like an LLM summary of a conversation.

2. No US educational institution should ever grade on a curve. Your job is not to compare students but to educate them. Grade curves hide the performance of the educators and process of education in actually improving the skills of students.

3. Both AI and the cognitive and emotional overload from social media taking away brain space may be to blame. Idea: let students report screen time statistics at the beginning of each semester and weekly or at the end. See if and how it correlates with academics.

There's often little discussion around incentives. Students cheat because grades are used as a major selection factor in university admissions. Maybe that should change.

Set a reasonable bar for grades or SAT scores and then use other criteria beyond that gate.