This is common mistake people make. You don't need an AI that can replace you, in order for you to be replaced by AI. You only need an AI that can replace one simple task you do in your work. If that task takes 10% of your work time, replacing that with an AI means that your employee can fire 10% of the employees.

I like to remind people that AlphaFold2 solved the protein folding problem. Extremely difficult problem in biology which no human team with the help of computers was able to solve in several decades they tried to. This is one kind of task that AI can do.

And this leads to second mistake people make. AlphaFold3 was made to help drug research. It is very likely that within few years, we have an AI that solves the problem of drug research. An AI that can make a drug for almost any decease, even those that are currently incurable. If we suddenly have thousands of new drugs that cure everything, how much of doctors and nurses we will need? So AI doesn't necessary need to do anything you do at work, it can simply solve the root problem, eliminating the need for your job completely. I think that this will be the major reason for people losing their jobs.

That is not the reason. There are two reasons:
1) Most people don't have knowledge, skills and motivation to improve their situation. Take food as a example. Almost everyone in western countries could eat more healthy, which would be an massive improvement to their health and life in general, but they don't do it, I think that at most 25% do this, but even this estimate is most likely too much. They could also invest money and become richer, but most won't do that either, I think that at most 20% do this.
2) Another reason is competition. For example if you want to produce content using AI, good luck finding your spot among millions of others who had the same idea. Or if you want to compete against Amazon or other big companies, they simply have much more everything than you, so it is pretty hard to compete.

- Some people look identical. If one person wants to commercialize their looks and other doesn't, who wins?
- If you make a copy of features, but change it a little, is it free game? How much you have to change?
- Why should a person have copyright to their looks anyway? Why can't everything just be public domain.
- What if you are a content creator and you makeup a random, but realistic look for human, but it ends up being copyrighted by someone. How can you even check it?
- What if you create a digital person and you have copyright for it, but then a new person borns and that person has those exact features. Is the newborn violating your copyright?

> Some people act dumb. Maybe humanity would be better served by spending our time and resources teaching them to act smarter.

Is that possible? If we had no limits with the money and we can spot a dumb kid at the very early age and we put the best teacher in the world to educate that dumb kid for years. Will that make that dumb kid a smart kid? Is there an upper limit of how smart the kid can become? How much effort would that require? What would be the sweet spot?

Reply to self. As mentioned in the link above, the open source version is not under the sanctions.

Lyon is 3rd largest city in France, population 520,774 (2022).

"Behind OnlyOffice hides the Company Ascensio System SIA, which have their Headquaters in Riga (Latvia) and London (Great Britain). The Company however is russian. Because of that OnlyOffice falls under the sanctions imposed by Germany together with the EU against Russia."
https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.en-zdv.uni-mainz.d...

So can France really use it?

Mathematical explanation:
https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3F...

Interesting example of how LLMs think (which explains partially why they fail):
https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3F...

> There is nothing wrong with AI as a computer science, but as deployed, it has brought nothing good to the world

AI has given as over 200 million folded proteins and millions of new materials
https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3F...
https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fdeepmind.google%2Fdiscov...

It has also advanced mathematics
https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fdeepmind.google%2Fdiscov...

Currently AI is trying to find cure for every known decease there is and most likely within few years it will.

Personally I have found several values for the LLM, for example:
1. Negative searches that were not possible with traditional search engines except in very rare scenarios. Like "Find me a molecule that has iron and oxygen, but not nitrogen."
2. Searching research papers with fuzzy words. It is very hard to find a research paper unless you know the exact words to search for it, but AI can translate your fuzzy wording into meaningful search results.
3. Testing out coding ideas. You can describe something you want and you get instantly code that creates the UI for you. The code might not be usable for the project as is, but you can test different ideas and refine your requirement specification easily with it.
4. When learning something new. Programming technology, human language etc. AI can explain reasonably well the basics of the subject. Sure, it makes mistakes, but I have never seen an official school book that didn't have any errors, so errors are everywhere. It is enough to get large part of things right when learning new stuff, you can refine that later when you know more.

Unlike the "AI ethic" claims, Google Deepmind has a definition for AGI, with progressive steps and a method to test whether you are going towards it or not:
https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Faibusiness.com%2Fml%2Fwhat...

The real news story is that CEO admits that AI will replace human workers. Before it has been all about "robots won't replace humans"

https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.digit.fyi%2Frobots-w...
https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.livemint.com%2FTechn...
https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.aa.com.tr%2Fen%2Fworld...

I have looked at the data for the past 15 years. Some work like healthcare and IT are going up and some work like manufacturing are going down. The total amount of work is declining steadily at the speed of about 0,4% of population losing a job every year. This is not actual job loss, as I am looking at hours of work done and estimating the amount of jobs based on the hours, because some jobs are short term, some have only few hours etc.

Problem is not about inventing new jobs. Problem is finding new customers. Currently new customers come from the middle class increasing by about 250 million per year. But that can continue for only about 20 years in the best case scenario. After that there will no longer be growth in consuming. After that, every increase in automation will not mean new jobs, it will mean just job loss.

And this is the absolutely best case scenario I can think of. I still think that most realistic scenario is that it will be just downhill from here on.

Why did they sign a deal instead of just switching to OpenJDK. We have a big project and we did the switch and there were zero problems with it. It is the exactly same product, but Oracle has some additional packages on top of it. If you don't use those rare packages, you can use OpenJDK just fine.

Personally I don't need to install solar panels, because so many have installed them already. For example today electricity cost is about minus 2 cents/kW on the cheapest time, because there is so much solar during those hours.

If I would install something, it would be batteries, but I have not done either. It is cheaper to just use electricity during the cheap times and avoid using it on the high price times. The electricity cost is about 10-20 dollars per month for me, which is insignificant with my income. I would not even have the need to use electricity during the cheap times, but it is like a hobby to save the planet.

Amount of green energy is growing at exponential speed. In 2024 green energy powered 40% of global electricity. Even US has increased clean energy production and switched from coal to natural gas, which is slightly better.

So situation is constantly improving at a rapid speed.