Many AI startups are using AI technology to find demand. To put it bluntly, they do n’t know what they are going to do, because AI technology does not create jobs, it just updates the tools. Example 1: Once humans cultivated by hoe, then by livestock, and then by machinery, tools became more and more efficient, but no matter how the tools changed, the demand for crops did not change. Example 2: Once humans exchanged items for goods, and later used paper money, and then used credit cards and PayPal. Although tools became more and more convenient, the demand for commodity exchange never changed. Example 3: Now that human demand for data has created new jobs that label data, it has allowed many people with disabilities to have a better income. Therefore, work changes with changes in demand, and work remains unchanged when demand is constant. However, the tools used in the work will be updated. The real danger in the future is that it will not be replaced by AI workers in the industry, but by workers in the industry who rely on a single tool. Even if the tool you master is the most advanced AI, it is equally dangerous, and even more dangerous. Because researchers are constantly updating the tools they invent, the more advanced tools are often invested in more research and development, so the fastest update rate. Today’s most advanced technology is likely to be completely replaced by the method in a paper tomorrow.
How AI is changing the industry
Thirty years ago, computers were also regarded as an emerging industry, but now people have long realized that computers are actually a tool to update the efficiency of all industries. All industries now use computers, and everyone uses computers. There are indeed a group of human beings replaced, but it is not that computers replace humans, but humans who use computers replace humans without computer assistance. As described in the following intercept: Humans who do not use artificial intelligence will be replaced. Now people ’s concerns about “AI instead of work” are exactly the same as people ’s concerns about “computer instead of work” in the beginning of the Internet era. But the fact is that the situation at the beginning of the era of artificial intelligence is also exactly the same as the situation at the beginning of the Internet era: instead of artificial intelligence replacing humans, humans are using artificial intelligence to replace humans that do not. This panic is different from the past because too many irresponsible media propaganda made people mistakenly believe that the current AI is autonomous.
Lee Sedol vs Alpha The human-machine war
Between Lee Sedol and Alpha can be said to be the tipping point of artificial intelligence. The media all interpreted the human-machine war as “Human lost to the machine”. But Dr. Aja Huang, who originally developed the Alpha Go, has been a genius of Go since he was a child (who once won the college cup Go tournament). And then joined the team with experts in many fields such as neuroscience, philosophy, and computer. Everyone is familiar with the rules of Go, and how to design the Alpha Go (more straightforward, speech recognition, computer vision, Natural language processing is also good, those AI experts themselves are experts in a certain field). This is where humans lost to the machine. This is clearly that Lee Sedol, who does not use AI, is “bullyed” by a team using AI. Can’t help but think of the poor candidates who did not bring a computer in the college entrance examination. Although AI dominates the chess world, it still does not affect people’s hobby of Go, just as we know that airplanes and cars have exploded into humans, but there are still Olympic Games. Instead of replacing the industry, AI has brought better teachers to the industry (Afa dog teaching system)
AI and Art Although the media and training institutions keep moving, the titles “AI replaces illustrator” and “AI replaces doctor” are outraged in all industries. But it is the AI technology that requires human control to generate pictures, create music, and transfer styles. Who will decide how to use the generated content at that time? Does it depend on the straight man’s aesthetic ability? Of course, it depends on the artist. To whom will these technologies developed by AI companies be sold in the future? Of course still an artist. Adobe’s software will be updated with more automatic tools every time (you can view the new features of AdobeMax Conference). Gradually including AI tools.
What if the pictures generated by the technology can perfectly meet the quality requirements in the future? This just means that artists no longer have to work hard to produce these AI-generated content, but can transfer their energy to other creative parts. Example: Current video editing requires creators to consume a lot of time to watch related videos, and now has AI automatic editing technology, which can allow creators to put time on the concept of the content, without having to mechanically repeat the viewing.
AI and medical, academic For this huge system of medical, AI can be empowered in many ways: virtual assistant, case analysis, medical imaging diagnosis, drug development, etc. Due to the inherent advantages of convolutional neural networks (CNN) for image processing, coupled with the support of data volume, AI has higher accuracy in medical imaging diagnosis. For example, the Chinese team published a paper in “cells”, saying that its main research and development of artificial intelligence (AI) tools for accurately diagnosing two major diseases of eye diseases and pneumonia, its overall accuracy reached 96.6%. But this is like the thermometer and MRI can accurately measure a certain index of the patient, but only improves the accuracy of a message. Do n’t forget that the doctor ’s job is much more than just looking at medical images. Diagnosis requires comprehensive judgment combined with many factors, as well as complex decisions such as prescribing medicine, surgery, and hospitalization. Look at how hard it is to train a qualified doctor to know that “AI replaces doctor” is really overhyped.
Don’t think that medical only imaging diagnosis but doctors who use AI are really different from doctors who do not use AI. We all know that this is an era of information explosion. There are countless papers published every year. No human can read all the literature in his lifetime to pick out the information that is useful for his research. I think everyone who does scientific research on this All have a deep experience. But if there is AI to help quickly read the relevant research from all the literature and improve the search efficiency, it will bring unprecedented help to the researchers’ work.
References:
https://en.wikipedia.org/wiki/Lee_Sedol
https://en.wikipedia.org/wiki/Aja Huang