The AI model appears to be a better and more reliable friend and verbal sparring partner than most other humans. Social interactions and consequences apply with humans, but not with AI. It never appears avengefull. The computer doesn't care and remains on-topic. Sober, drunk, leftish, rightish talks. It will have an opinion but refrains from judging me. What is weird about it, apart AI slowly taking over the world by using humans by using suggestion to influence us (me).

The real question would be: why doesn't internet provide me with a group of likeminded to philosophize. Why does AI do a better job than real humans, which tend to seek conflict. AI at least tries to please me. We humans build an internet to collaborate and all we do is disagree.

Given screentime, i'd say Android is my main 'desktop OS'. Yes, one should call it a ''coffeetable OS''.

If I was Satoshi i would never admit committing such crime against the planet.

Dear Windows.

Just do what i usually do, so i can take a break.

Sorry, but Google is just as annoying when you visit them. Always pushing their browser. It's childish behavior from big tech and users suffer. I already have a browser, thanks.

Each pound of ice you make in your fridge, costs about 3 pounds of ice from Antartica to make.

#$!

The emergence is, that a set of simple rules result in complex behavior. A nice example is Conway's game of life. Simple rules, complex behavior. We consider current GPT models to be intelligent, and so i say we have intelligence as emergent behavior.

And yes, it is just a model. Cause that is what computers do. Meanwhile it's pretty useful. Once we toss enough hardware and electricity to the problem, we can even chat with it and it can do things for us.

By calling it 'just a statistical model' you ignore the emergent property of intelligence. It doesn't really matter if the computer understands what it says, as long as humans do. And when it comes to AI, i'd say, as long as it's useful, use it. When i use a hammer i never worried if it's internal atomic structure and quantum spins bother me beating it on something. It is an irrelevant question.

Now as for the writers, i understand them. I also think AI at this moment is really not that good at story telling. Yes, it can spit out information. But there's a reason we pay professional writers to write quality stories, instead of the youngest clerk. But, for programmers, they tend to view it in a different light. Programmers don't care or are not afraid of AI taking over their job. They are just happy with anything they don't have to type, so they can focus on other things, like more features, the overall structure, gluing stuff together, fixing bugs etc. Also, it is quite obvious current AI is really not that good at writing (complex) code (just as writing quality stories). And it does make other kinds of errors. Human errors are often typo's. AI errors are often wrong order, failed assumptions, or not understanding the bigger picture how things interact, let alone screens and user input. Meanwhile, it is a very valuable tool that saves the programmer a lot of time and work, so that he can be more productive. But you still have to be a good programmer to use the tool properly. And long paragraph short, i am not seeing why the situation is significantly different for writers, or any other profession.

AI will only replace you if you suck at you suck at your job. Because someone else that is better at their job will do your job as well, since they can be more productive. Also, last i know the world is still screaming for more programmers and more technical people. Or lawyers. Or artists. Or other professions. Yes, AI may develop to the point where it actually does better (than humans), by more feed back, more real world integration etc. But, we are at least a decade away from that, if it ever comes. My hand is not a hammer or plier, it uses fingers which are not made of metal. Does that make the hammer superior to my fist? Of course not.

Really missing that at the moment.

Apart 2 stories (yesterday and today) on /., i have no indication that it happened. Also, guts tell me FB implemented some auto-signup whenever you used one of their platforms. It sort-of contradicts my previous statement, that outside /., i did not notice its existence. Going by statistics, numbers don't align to me and seem off, one way or the other.

Ok, as paying open-AI user, to reap the benefits of GPT-4, first question that comes to mind is, why would i pay?

Indeed. It won't even run on my phone, despite all specs. And by your description it is the same as with their other games, to much bloat that actually hinders gameplay instead of supporting it, and crappy compatibility on older phones for some games.

Also, they try to create revenue but they don't actually have something to sell. On one hand i'm happy that they didn't join the pay2win hype, and if they did i'd probably stop playing. But they also don't seem to listen to, or support their user and fan base a lot. And their key technology/data collection, as in the wayfarer systems, sucks so bad that most users get discouraged contributing, as they feel penalized and don't receive proper feedback. So they have a huge userbase totally willing to make the quality of the database top notch, but they deliver a year-2000-style UI and a raise a thick middlefinger to those contributing.

To sum it up, i am not surprised they are not able to make more revenue. They have some good ideas but they seem to focus on the wrong things and set the wrong priorities. And instead of launching new games, they should seek to keep their current players by improving the games they have. Also, just cancelling games seem like a waste of effort to me. It's a shame, it's a promising company really but ran by monkeys. Or actually, monkeys would do it better.

The field of AI could right now already be useful for translations. Quite often some information is only available in an English, German, Spanish or French wiki entry. And yes, translate tools exist but may not be as good as a quality controlled AI translation.

Having said that, my main worry would be the future feedback loop, as AI generated information will be used to train future models. Its highly recommended to tag any page or paragraph that was generated by AI as having an AI author. Apart that, i do not see real issues if it helps humans to speed up generating quality content.

Netherlands has one commercial nuclear power plant left. It's actually build right next to the sea, in a polder, in the town of Borsele. Apparently it can withstand sea levels of +9 meter (about 37 feet) above sea level. And even in case of a flood, the reactor is supposed to be safe. The risk of a tsunami in this region of the world is considered nihil, though storms might wreck havoc. Yet, a disaster like in Japan is highly unlikely.

Personally, i'd rather live next door a power plant (they decommissioned a nearby powerplant after public protests) than any other power plant. Some might even include windmills, although so far they didn't bother me. France, Germany to a lesser degree, Belgium and eastern Europe countries have plenty of nuclear plants. I'm fine and wish we had build way more so or summers wouldn't get so hot.