Ma vs. Musk – is AI boom or doom?

Jack Ma and Elon Musk recently debated the future of AI, with one believing AI will bring humans more freedom from the menial tasks, and another seeing in AI the end of human beings in the current shape.

Jack Ma, the founder and former head of Alibaba and now the Co-Chair of the UN High-Level Panel on Digital Cooperation, and Elon Musk, the founder of Tesla, SpaceX and other ventures, took to the stage at the World Artificial Intelligence Conference (WAIC), currently being held in Shanghai, China, to debate the virtue and vice of AI.

The dialogue, unmoderated, sometimes felt awkward, when the two looked to struggle to find a common anchor point. (The full dialogue is embedded at the bottom of this report.) But there were also agreements occasionally, for example both agreed AI will displace many jobs. However, the two entrepreneurs took very different views on the role AI can ultimately play, especially when it comes to its impact on the future of mankind. Ma took a rather utopian view, claiming AI can help human beings understand and take care of ourselves. He conceded that lots of the jobs many people are doing now will be lost to AI, but he saw that as a positive thing, because “I think people should work three days a week, four hours a day.” There were also a few throwaway claims, like “In the artificial intelligence period, people can live 120 years,” therefore “we need artificial intelligence for the robots to take care of the old guys.”

Musk took a much darker view on AI. He believed the ascendency of AI, with its much higher “bandwidth” than human brains (“a few hundred bits per second, basically, maybe a few kilobits per second, if you’re going to be generous), will render human jobs “pointless” and ultimately take over everything. “Probably the last job that will remain will be writing AI, and then eventually, the AI will just write its own software,” Musk predicted.

It has to be pointed out that both men have invested interest in the topic and the viewpoints they took reflected their interests. Ma, despite that he had stepped down from the CEO’s position, would not be able to dissociate him from Alibaba. His quip at the beginning of the dialogue that “I would like AI to mean Alibaba Intelligence” certainly did not help the perception that he is detached from the business. Alibaba is one of the world’s heaviest user of AI both in e-commerce and increasingly in its cloud computing business – the company acquired Whale Cloud from ZTE to dovetail with its own Alibaba Cloud to serve different clients. Additionally, AI was supposed to play an important role in making lending decisions by Ant Financial, an Alibaba affiliated company, but it was reported earlier the system has not been reliable enough.

Musk’s interest in AI, and its link to the views expressed at the conference, is more complex. He founded OpenAI, a research company, but had decided “to part ways on good terms” with it in 2016. Tesla, and the autonomous car market in general, will increasingly use AI. But more recently he has been directly involved in brain-machine interface (BMI) with his new venture NeuraLink. In July the company applied to US regulators to start trialling its probe device on humans. The flexible threads, thinner than a human hair but connected with over 3,000 electrodes and able to monitor the activity of 1,000 neurons, could connect specific areas of the brain to computers. The first target was to provide AI support to paralyzed patients.

So, there was little surprise when Musk advocated connecting the low bandwidth human brains to the computers that “can easily communicate at a terabit level”, so that human beings could “go along for the ride with AI”, or what he called “symbiosis with artificial intelligence” when he was introducing the new NeuraLink technology earlier. He saw in AI a future where AI will be able to “completely simulate a person in every way possible.” He even went philosophical at WAIC by saying “there’s a strong argument, we’re in the simulation right now.”

(This “strong argument” is not actually new. Chuang-tzu, one of the Taoism masters from 4th century BC, famous stated “Now I do not know whether I was then a man dreaming I was a butterfly, or whether I am now a butterfly, dreaming I am a man”. There are at least two counterarguments to refute this speculation. One is it cannot go through the Popperian test, that is the argument cannot be falsified. Another is there are simpler answers to address the nature of the question, with lower level of entropy, therefore should always be preferred. Both arguments have been extensively explored by Prof David Deutsch, the quantum physicist, in his 1997 book “The Fabric of Reality”.)

Incidentally, Elon Musk recently endorsed Andrew Yang for the 2020 presidential election. Yang, the entrepreneur-turned candidate, champions universal basic income, arguing such a measure would provide the basic safety in the face of massive job losses to AI. According to the research quoted by Yang, among the most vulnerable groups would be the truck drivers and retail cashiers which are generating the biggest number of jobs in America nowadays. It would be very hard to retrain these people quick enough to handle AI-powered positions. In that sense, Jack Ma’s claims that “don’t worry about the machines” and “we will have jobs” may be too optimistic.