#4 Hurdles to go beyond Artificial Intelligence (Ai) to be usable

In daily life, there is no day without " Ai " word. Future evolution is almost certainly no mistake, but still, it seems there is still something necessary to achieve further leap.

IBM is a representative of companies that are focused on Ai, such as Watson, which is a cognitive solution. However, the Web media infinity, which is operated by the company, said that As Ai is acknowledged by society, Four conditions was published in the article.

The first thing mentioned was " small data ". As you know, Ai is good at analyzing a huge amount of data and offering an optimal solution, but in the future, it seems that it is expected that it will be able to assist humans with fewer data.

The second is about the evolution of " unsupervised learning ". Machine learning is basically a mechanism to teach the correct answer to a certain question and to learn it by repeating it many times, this is called "supervised learning". On the contrary, it is called "unsupervised learning" which does not teach the correct answer to the computer and advances the answer to be derived also undetermined. It is necessary for the computer itself to infer the purpose and correct answer, and very advanced technology is required, but it seems that it is expected as a proof of the further evolution of Ai by that much.

The third article is an article on "Recommendation on improving hardware " and the fourth on " Recommendation to raise the reliability of Ai " this time. If you are interested in evolving Ai, please enjoy continuing from infinity.

---