People start using AI as a tool to supercharge their trolling? Don't worry, it won't put any trolls out of business. It just allows each troll to accomplish five times more!

I'm skeptical that this is actually an AI and not in fact a person trolling. That sounds exactly like what someone would do as a joke. One of the stranger elements of AI literacy these days is remembering that things claiming to be AI generated sometimes aren't. Often it's really a person just pretending.

This is so dishonest. Everyone pays for OpenAI's products. They force you to pay in indirect ways you can't avoid. If your electric rates have gone up, thank OpenAI. If you have to pay more for a new computer because memory has gotten so expensive, thank OpenAI. Not to mention the huge amounts of CO2 being dumped into the atmosphere to power their data centers. Every single person subsidizes their products, whether they use those products or not, and then OpenAI pretends to be giving them away for free.

That's why the propaganda works. If you tried to tell people in California, "Look out for the scary Japanese who are invading," most of them would laugh at you. Because those scary Japanese are their friends and classmates and coworkers who they've known for years. On the other hand, if you tried to tell them, "Look out for the scary Martians who are invading," they would just get confused and wonder what Martians you were talking about.

The South is currently balanced between these states. There are enough immigrants around to be visible, but they're new enough to still be scary. They aren't the people you grew up with or work with every day. That will change over the next few generations, as it changed in California over the last few.

Maybe, but I think a different explanation is more likely. It's related to the changes in immigration over the last few decades. This article has lots of statistics about it. This sentence is the key:

What you're seeing is people who aren't used to immigrants suddenly having lots of immigrants in their communities. The reaction in the South today is very similar to the reaction in other parts of the country a century or more ago. If you were Irish in New York or Japanese in California in the late 1800s, you faced incredible hostility and discrimination. With time, those regions came to accept immigrants and view them positively, but that wasn't always the case. Today the South is getting lots of immigrants for the first time, and people are reacting in the same way.

I predict his prediction will come true right around the same time he arrives on Mars. I'm really looking forward to one of those.

This shows why the AI spending boom isn't sustainable. One one hand, the only barrier to entry is willingness to spend insane amounts of money. Which certainly is a barrier, but there are enough large companies willing to do it that the market has become very competitive.

And on the other hand, there's no ceiling on how much you have to spend. The only requirement is, "More than your competitors." It's an arms race. Every company has to keep spending more and more to stay competitive. The spending grows without limit and the benefit they get from the spending doesn't.

The best case for them is that it becomes a war of attrition. Companies drop out as they decide they can't afford the spending. Eventually it narrows to just a few companies, and they agree to stop spending so much and share the market.

The worst case for them is that it all gets disrupted by other companies (e.g. DeepSeek) that figure out how to compete without spending so much, and their massive investment becomes worthless.

If you consider everyone in the business world, I think the rate is much lower than that. But if you consider only people in senior management positions, 3% sounds about right to me. That's based on people I personally worked with who later did or were found to have done illegal things they could have been sued or prosecuted for.

But none of them actually was sued or prosecuted. There would have been nothing to gain from it. Once people figured out what they were doing, the response was to show them the door and try never to see them again.

The same is likely true of the Forbes list. If 3% get charged with crimes, the number committing crimes and not getting charged is probably much higher.

That's why they're focusing on middle mile, not last mile. Last mile is hard. Your truck needs to go anywhere, and you don't know if there will be someone there to unload it at the end. Middle mile between warehouses and stores is a lot easier. It only needs to drive fixed routes that you assign in advance, and you know there will be someone to unload it at the end.

Why not focus on long-haul highway driving? Because other companies are focusing on that. This is a different market that requires a different product.

In a post that it will soon be illegal for anyone under 15 to read.

Leaders need to practice what they preach. If you're going to cut teens off from social media, don't put important information on social media where they can't access it.

The post you're replying to said Christian Nationalist, not Christian. They're not the same thing. See the Wikipedia article which begins, "Christian nationalism asserts that the United States is a country founded by and for Christians."

That's not how venture funds work. The company managing the fund (Mozilla in this case) doesn't invest their own money. They find investors who want to participate, and get a share in all the companies they invest in as payment for organizing everything.

It's not a moving target at all. The term was coined in 1955, and its accepted definition hasn't changed the slightest bit in the 70 years since. It comes from the proposal for the Dartmouth Summer Research Project on Artificial Intelligence. Here is how they defined it.

That is still exactly what practitioners in the field take it to mean. If you come across other definitions, that's people making up their own definition because they don't know how the field has defined it for 70 years.

Two important points about this definition. First, artificial intelligence is defined as intelligent behavior. If a machine behaves intelligently than it is intelligent. How that behavior is implemented doesn't matter. Second, if a machine can do anything a human can do then by definition it's as intelligent as the human.

These ideas were strongly influenced by Turing's "imitation game" (now known as the Turing test) that he published five years earlier.

I challenge anyone to come up with a definition of intelligence that 1) is clearly satisfied by humans, 2) is clearly not satisfied by current AI, and 3) isn't totally contrived (like defining intelligence as "the ability of humans to...").

Here are some common definitions of intelligence.

"The ability to take in information and make decisions based on it." Computers do that.

"The ability to solve problems." Computers do that.

"The ability to predict the consequences of your actions." Computers do that.

"The ability to understand complex subjects." Humans often speak or reason about things without really understanding them. For examples, see the comments thread on any slashdot story. Besides, "understand" is another word that doesn't have a clear definition.

If you believe AI isn't intelligent then how do you define intelligence? Are you absolutely certain AI doesn't satisfy your definition? Are you absolutely certain that humans do? And if you answered yes to both, is it a contrived definition that doesn't match how most people use the word?

I hope you meant that as a joke! If not, it's wrong in so many different ways.

The difference between data and anecdotes is that data is sampled from a probability distribution. By looking at the samples, you can learn about the distribution they're drawn from. That lets you make predictions about future observations drawn from the same distribution.

Anecdotes are chosen individually. They don't reflect any probability distribution. They don't let you predict future observations.