Kai-Fu Lee’s visions from the future
[ad_1]
Twenty years ago the panorama of technological innovation was radically different from today: there were no Facebook, Twitter or even smartphones. Virtual reality had all but been abandoned after the experiments of the 1990s. The deep learning that powers artificial intelligence today was little more than a chimera, studied almost secretly by a small niche of scientists. In two decades the world has transformed: social networks and smartphones have changed society forever, virtual reality – despite some difficulties – is experiencing a second youth, while artificial intelligence has revealed itself to be a truly transformative technology, exploited every day coveted and which continues to surprise us with its continuous evolutions.
What will the tech world be like in another twenty years? They tried to answer this question AI 2041: scenarios from the future of artificial intelligence (ed. Luiss) the Taiwanese computer scientist, essayist and investor Kai Fu Leea leading artificial intelligence expert, and Chen Qiufan, one of China’s best-known science fiction authors.
The ten short stories written by Qiufan, set in 2041 in the four corners of the planet, are accompanied by short essays by Kai-Fu Lee, who address and explore the potential and risks of the technologies featured in the stories (mixed reality, quantum computers, autonomous cars, AI , deepfake…), as well as realistically imagining what level of maturity and impact on society they will have reached in twenty years. We interview Kai-Fu Lee, connected via Zoom from Taiwan.
Artificial intelligence
ChatGPT: a weapon of mass approval?
by Guido Scorza*
Although the essays and stories have been created in synergy, sometimes there is a slight contrast between the two parts of the book: if there is a dystopian vein in the stories, sometimes even accentuated (as in the case of the compelling “Quantum Genocide”), in the non-fiction part, the optimism and enthusiasm that has always characterized it prevails. Which of the two visions will prevail in our future?
“It’s understandable that technology raises concerns. But we should take a step back and look at things in a historical light. The automobile was also considered a scary technology, as were electricity and personal computers. If we look at the great innovations technologies, have always seemed scary at first and there have always been people who have asked to stop everything. Sure, in the short term they can create problems, but in the long run, all these innovations have brought more benefits than problems. other things, as I explain in the book, technology is also able to find solutions to the problems it creates: antivirus for computer viruses, life-saver in cases of electrocution and one day we will also find solutions for fake news or intelligence errors Of course, it is not guaranteed that this will happen, especially in the case of AI: ChatGPT for example opens a Pandora’s box in terms of possible ab uses and dissemination of false content. In general, though, we’ve always gotten more benefits than problems from technology.”
Just ChatGPT has been the undisputed protagonist of the last few months, a tool that – in addition to the risks you mentioned – raises fears that one day it could replace man not only in the most banal and repetitive tasks, but also in fields that require creativity or even sensitivity. It’s a scenario imagined in AI 2041 in the short story “The Job Savior,” where you imagine AIs interviewing and selecting staff without candidates even noticing. How credible is such a scenario?
“The HR sector is certainly among the most difficult to replace, because it requires empathy, connection, trust: all things that the machine is not able to have. Personally I am amazed by the progress of ChatGPT; having said that, it is clear that is still ready for the big leap: it cannot, for example, write news or be used as a search engine, because it cannot recognize what is true or false.However, it can be a fantastic assistant, who has all the knowledge of the world at its disposal and he is able to retrieve information. Of course, he is an assistant who has the bad quality of inventing things from time to time, but the important thing is that there is a human being who selects and filters”.
In short, the road to switch from ChatGPT to the so-called general artificial intelligences – of a level equal to or higher than the human one – is still very long.
“ChatGPT is trained to statistically predict the words that can complete a sentence or answer a question. It seems to converse in a sensible way and this is already miraculous. What we now have to do is therefore train these tools to abstract, to remember, to true understanding and cognition. It’s just the beginning and it’s still going to be a long time, but I’m personally excited.”
Ideas
Bill Gates’ predictions about the future of humanity in the age of artificial intelligence
by Archangel Rociola
While deep learning has already demonstrated its enormous potential, there are applications that seem to have reached a stalemate, such as autonomous cars. At the same time we are still in the embryonic stage of another potentially revolutionary technology such as that of quantum computers. Both are the protagonists of the stories of AI 2041: how many possibilities are there really that within twenty years they will become consolidated realities?
“I think all the technologies in the book will come true and my predictions are actually quite conservative. The only one I’m not 100% sure about is quantum computers, which may take longer. As for the autonomous cars I have no doubts, also because they are gradually improving and learning to manage increasingly complex scenarios.Be careful though: many American companies, such as Waymo for example, are trying to immediately create the perfect product to be used in all scenarios. admirable ambition, but I prefer the Chinese approach: produce something simple that works in defined environments and can already have commercial applications. You can then continue to build on this starting point”.
AI 2041 was written when Mark Zuckerberg had not yet announced his metaverse project, but in the book there is still a virtual, immersive, open environment in which people – in a world where Covid has become a threatening constant presence – have transferred a large part of their lives. Little wonder, given that you are one of the pioneers of virtual reality, on which he was already working in the nineties. What do you think of the metaverse?
“I think it will still take a long time for a vision of this type to take hold. Today, behind the metaverse there is mostly the hype generated by companies that are too optimistic about the possibility that this project will materialize and spread. It is not so easy to create an immersive environment, in 3D and of quality using comfortable viewers, which do not require too much energy, do not cost too much, etc. I also do not believe in the metaverse for social interactions, for commerce and other serious applications of this type.Over time it could happen, but it is not natural for us to make friends in virtual reality or to buy real estate. Unless this happens in gaming, the area in which I see the greatest potential together with films and entertainment. Before it spreads, it will still take three or five years. And then we’ll see how it evolves.”
Religion and artificial intelligence
If Pope Francis talks about neural networks with Jews and Muslims
by Jaime D’Alessandro
[ad_2]
Source link
