Next up: email address auto-completion powered by Gemini!

The real reason the EU is embracing open source is because it doesn't want to be dependent on U.S. tech firms (understandable). But the real reason that they are dependent is b/c they don't have a strong domestic tech industry themselves. Shifting to open source will not solve the underlying issue.

Most companies record badge-ins, but not badge-outs for fire-safety reasons. To measure time spent at the office, they usually link to wifi or other network activity. The easy solution is to go to work, plug in your laptop, and leave.

Anecdotally, Slashdot comment counts are also down from what I remember them being. Same on other forums like hackernews. Stackoverflow had its problems, but AI slop lowered the incentive for anyone to contribute online.

> Human mistakes are of an entirely different nature and quality than AI 'mistakes'. A human won't accidentally make up facts, cases, or sources. Actually, yes they do sometimes. The big difference between a human and an AI is that humans are accountable. If a human messes up, they get fired and there is remediation. If an AI messes up, we say the technology is maturing, good luck.

Section 230 of the communications decency act gives online service providers the freedom to publish anything they want. Repeal it, and this problem goes away.

The real story here is that an individual went up against a giant company on his own on a relatively complex contractual legal issue and won without the help of lawyers. I'd bet my lunch he used AI to help organize and draft his legal filings, and I'd bet further this case would never have been brought had it not been for the help of AI.

This is the promise of AI, helping yourself through complex issues to get a real result.

I'm surprised people are actually listening to this dribble instead of just looking up the wikipedia article. If automated content can actually engage humans, we have no chance.

The UN is saying that for every $100 of economic growth (like an buying Air-Conditioner), there $40 of environmental harm. That leads to ridiculous conclusions: no one can do anything because it marginally can harm others (sorry, freedom); and poor countries that need fossil fuels to develop are expected to technologically "leapfrog" advanced nations and magically start out with an advanced economy (sorry, Africa). The UN admits that the changes they recommend don't have benefits until 2050. Overall, it seems that they are advocating for more harm, not less.

I get some people are afraid the world is going to shit, but policing thought isn't the solution. The solution is for everyone to distrust information and then spend time to dig deep. A fact isn't truth, it's information. How much information do you need to understand the problem varies depending on the topic

the downside of thought police and "fact checkers" is they fool themselves into thinking they know better. Everyone is different and keeping information from people is patronizing bullshit. Instead of shutting down people, the only real way to move forward is for everyone to talk to each other. Of course most people suck at listening and are locked into their own belief system, but policing thought just makes it worse.

Would need to replace 10 to make it worthwhile.

of course Ackman is bitching and acting like a bully. Vampire capitalist don't actually give a shit about other people, they care about money, power, money, power and money. The more important question is this 'why do people in the US keep listening to dick head billionaires who have zero moral authority?"

I love how it's totally cool for OpenAI to use unlicensed data from the internet and not credit open source projects to build their own model. As soon as someone else does that to them, it's not right. What a load of hypocritical shit.

I contribute to open source and I don't mind OpenAI using it to train models. But just like a human, you better damn well cite and give credit to the projects.

you realize Intel is using TSMC now and they are still struggling after over a decade of bad CEO. The other big issue is the culture in the US. North american kids don't want to put in brutal hours in a chip fab. Who is going to run those fabs?

it's not that simple. The models memorize a significant amount of the data it was trained on. So it is violating copyright since it has a copy of the original artwork in the neural network weights. To get around this, the technology has to progress to the point where engineers can clearly prove the model didn't memorize the training data. We still can't interpret the weights. Don't take my word, go search arxiv.org for papers.