Do you know, what unites planet hunters, stock marketers, and cancer doctors?
All of those people are working in areas where statistics are important things. By following statistics planet hunters can search exoplanets. The changes in the star's brightness. When the planet travels between them and Earth tells that there is an exoplanet. But noticing those changes requires long-term intensive data collecting. The planets that are orbiting very close to some red dwarf cause very short periods and easily seen changes in the star's brightness.
But if we want to search Earth-type planets that are orbiting yellow stars. We cannot use that method. The orbiting time of those planets is so long, that it requires years of intensive work. And those telescopes have limited observation time. The thing is that AI can measure the brightness of many stars at the same time. The AI can also search for differences in brightness from multiple stars. And that allows it to collect information extensively.
Extensively collected data can make a person a very successful stock marketeer. The problem is that the person who is working in that business must follow the advancements of the stocks in companies. But the problem is that intensive following and statistics collections require time. And if the person just follows the wrong companies that is the loss.
The AI can follow multiple companies at the same time. And because the data is collected extensively, the success is not depending on one company. The investor can select from multiple companies. And because that person can use a network of companies one mistake doesn't cost so much like a case where all money is put into one company.
Chat GPT makes many things better than humans. And it beats humans in things like the stock trade and the AI makes the diagnosis better than doctors. The reason for that is the AI notices all variables. Another reason for that thing is that success in stock markets is based on the long-term experience of how the single target's course behaves. Experience in stock marketing is not very important. The important thing is how accurate statistics the person has. The problem is that something can happen when that person sleeps.
But collecting statistics requires time. The problem is that one person can follow only a limited number of targets. And if the target is wrong, that thing means, the money will be lost, or profits are smaller than they could. The AI can follow the statistics of a large number of companies. It never sleeps. And that thing makes it an excellent stock trader.
That means the person who is making a stock exchange should follow the object 24h per day, and then make a decision does the marketeer buy or sell. Everything is based on the ability to collect data from multiple targets and keep statistics about behave of those targets' courses. The thing is that AI can make very accurate statistics from many targets. And it can use multiple sources.
When the AI searches for cancer from X-ray images it just searches for differences in the brightness of certain parts of the image. If that AI has a complete database of the X-ray (or MRI etc.) images in its memories it can compile the changes of brightness in areas that are seen in the images, and then it can make a diagnosis. It can also collect information that it gets from blood samples.
https://scitechdaily.com/new-harvard-developed-ai-predicts-future-pancreatic-cancer-up-to-three-years-before-diagnosis/
Comments
Post a Comment