Ray Kurzweil, worldwide authority on AI research, talks about optimistic prospect that 'progress of AI will progress smoothly'

Ray Kurzweil , an American inventor and businessman, is a world authority on artificial intelligence and is known for his work on " technical singularities (syngularity) ". At RAAD 2018, a technology conference on rejuvenation and lifetime extension held in 2018, Mr. Kurzweil talked about AI's progress and future prospects.

Ray Kurzweil: 'AI is Still on Course to Outpace Human Intelligence' - Gray Scott
https://www.grayscott.com/seriouswonder-//ray-kurzweil-ai-is-still-on-course-to-outpace-human-intelligence

This movie was filmed by Mr. Gennady Stolyarov, an American transhumanist party chairman, serving as an interviewer and hitting Mr. Kurzweil with AI related questions.

Gennady Stolyarov II Interviews Ray Kurzweil at RAAD Fest 2018


When Mr. Kurzweil appears, applause is sent from the venue. Male on the left side is Kurzweil.

Asking the interviewer, "What do you think of AI robots like Sofia that appeared at this venue earlier?" Mr. Kurzweil said, "I got a very strong impression, the time required for response is short," I mentioned.

Mr. Kurzweil made predictions of "popularization of the Internet" and "computer winning in chess matches" from about 1990, and they made a lot of efforts in a time lag. Also, "One invention is related to another invention, and one invention shortens the time until the next important invention, whereby progress in science and technology progresses exponentially rather than linear graph We announced the law of acceleration of harvesting called "a lot of discussion." And in 2005 he developed a stepped theory on singularity, and he also spread the concept of singularity in the world.

In recent years, it has been confirmed that AlphaGo (Alpha Go) developed by Deep Mind participating in Google has won the top players, and AI has begun to exceed human intelligence in specific tasks.

"The game of AI can not win against Lee Se-dul but I can not win" The strongest players in China will be able to beat Alpha Go - GIGAZINE

Mr. Kurzweil predicted that the intelligence of computers would exceed humans around 2029 and that in 2045 the ability of computers exceeded human imagination. In this interview held in 2018, Mr. Kurzweil's argument has not changed, "AI will eventually outnumber human's intelligence".

I touched on an attempt by the interviewer to apply the AI system " Watson " developed by IBM to medical diagnosis and asked, "Does AI not learn incorrectly like a cognitive bias that humans have?" ... ...

Mr. Kurzweil replied that progress in AI is remarkable and bias of cognitive bias of AI which was once worried can be dealt with. It is possible for AI to distinguish between fake news and truth news, hate speech and non-speech, and not to learn problematic data.

Due to the positive tone of AI and science and technology, Mr. Kurzweil is sometimes called "technology ultra optimist". However, in the interview, there was no change in optimistic thinking towards its own AI, and developed the argument that "Entrepreneurs should be optimistic about science and technology."

Matters concerning privacy and human dignity are high in priority, not only engineers but also politicians and others can join together to create safety guidelines that respond to advances in science and technology, Kurzweil says.

There are movies and novels that "human beings fight against the world of disciplines managed by AI", but there is no centralized AI to manage people at the moment. Rather, each AI is mounted on smartphones in people 's pockets, and it can be said that AI is in a very decentralized state. Of course there are concerns about gigantic companies such as Amazon, Facebook and Google, but at the same time the move to democratize technology is also becoming active, and Mr. Kurzweil argues that it will not be a threat.