Sam Altman, President of Y Combinator
Patrick T. Fallon | Bloomberg | Bloomberg | Getty Images
Artificial intelligence companies could become so powerful and wealthy that they could offer a universal basic income to every man, woman, and child on earth.
This is how some members of the AI community interpreted a long blog post by Sam Altman, CEO of the OpenAI research laboratory, published earlier this month.
In just 10 years, AI could generate enough wealth to pay each adult $ 13,500 a year, Altman said in his 2,933-word play titled “Moore’s Law For Everything.”
“Working at OpenAI reminds me every day of the scale of socio-economic change, which is happening sooner than most people believe,” said Altman, former president of the acclaimed startup accelerator Y-Combinator earlier this month. “Software that can think and learn will always do more of the work that people are doing now.”
Critics, however, are concerned that Altman’s views could do more harm than good and that he is misleading the public as to where the AI is going.
Glen Weyl, an economist and principal researcher at Microsoft Research, wrote on Twitter: “This wonderfully embodies the AI ideology that I believe is the most dangerous force in the world today.”
An industry source asking to remain anonymous because of the nature of the discussion told CNBC that Altman “envisions a world where he and his colleagues will become so immensely powerful from the AI CEO that they will make any non-AI- Corporations (employing employees) are pulling corporations and every American worker into unemployment. So powerful that a percentage of OpenAI (and its colleagues) income could fund UBI for every citizen of America. “
Altman will be able to “get away with it,” the source said because “politicians are drawn to its immense tax revenue and the popularity that paying their constituents’ salaries (UBI) will give them. But that’s an illusion. Sam is no different from other capitalists trying to convince the government to allow an oligarchy. ”
One of the main focuses of the essay is the requirement to tax capital – business and land – instead of labor. That is where the UBI money would come from.
“We could do something like the American Equity Fund,” Altman wrote. “The American Equity Fund would be by taxing companies above a certain valuation of 2.5% of their market value per annum, payable in shares transferred to the fund, and by taxing 2.5% of the value of all privately owned real estate payable in dollars. “
He added, “All citizens over 18 would receive an annual distribution in dollars and company shares in their accounts. People would be mandated to use the money as they needed or wanted – better education, health care, housing, business, whatever whatever. “
Altman said every citizen would get more money from the fund every year, provided the country continued to do better.
“Every citizen would therefore increasingly participate in the freedoms, powers, autonomies and opportunities that come with economic self-determination,” he said. “Poverty would be greatly reduced and many more people would have a chance at the life they want.”
Matt Clifford, co-founder of start-up construction company Entrepreneur First, wrote in his newsletter “Thoughts in Between”: “I don’t think there is anything intellectually radical here … these ideas have been around for a long time – but it’s fascinating to show how mainstream these previous fringe ideas have become among the tech elites. “
Meanwhile, Matt Prewitt, president of RadicalxChange, a nonprofit movement for the next generation political economy, told CNBC, “The play sells a vision of the future that makes and would make it far too easy for our future overlords.” likely to create some kind of peasant class that encompasses most of society. “
He added, “I can imagine worse futures – but this is the wrong direction in which to steer our imaginations. By instead focusing on ensuring and enabling deeper and broader participation in political and economic life , I think we can get far better results. “
Richard Miller, founder of technology consultancy Miller-Klein Associates, told CNBC that Altman’s post was “messed up”, adding that “the model is unrestrained capitalism.”
Michael Jordan, an academic at the University of California at Berkeley, told CNBC the blog post was too far removed from intellectually sound reasons, either technological or economic, that he would rather not comment on.
In Altman’s defense, he wrote on his blog that the idea should be little more than a “conversation starter”. Altman didn’t immediately respond to a CNBC request for an interview.
An OpenAI spokesperson encouraged people to read the essay for themselves.
Not everyone disagreed with Altman. “I like the proposed wealth tax strategies,” Deloitte employee Janine Moir wrote on Twitter.
OpenAI was founded in 2015 by a group of entrepreneurs, including Elon Musk, in San Francisco and is considered one of the best AI laboratories in the world alongside Facebook AI Research and DeepMind, which was acquired by Google in 2014.
The $ 1 billion research lab, backed by Microsoft in July 2019, is best known for creating an AI image generator called Dall-E and an AI text generator called GPT-3. It has also developed agents that can beat the best people at games like Dota 2. But it’s far from developing the AI technology Altman described, experts told CNBC.
Daron Acemoglu, an economist at MIT, told CNBC, “There is an incredible amount of false optimism about what AI can do.”
According to Acemoglu, algorithms are good at performing some “very, very tight tasks,” and they can sometimes help companies cut costs or improve a product.
“But they are not that revolutionary, and there is no evidence that any of this will be revolutionary,” he said, adding that AI leaders are “getting lyrical about what AI is already doing and how it is revolutionizing things”.
In terms of the measures that are standard for economic success, like total factor productivity growth or output per worker, many sectors have had their worst time in about 100 years, Acemoglu said. “It doesn’t compare to previous periods of rapid technological advancement,” he said.
“If you look at the 1950s and 1960s, the TFP (Total Factor Productivity) growth rate was about 3% per year,” said Acemoglu. “Today it’s about 0.5%. That means you are losing about one and a half percent GDP (Gross Domestic Product) growth every year, which is a really huge, huge, huge slowdown in productivity. This is completely at odds with this notion that we only get a tremendous amount of benefits (from AI). “
Technology evangelists say AI will change the world for years, speculating that “artificial general intelligence” and “super-intelligence” are not far away.
AGI is the hypothetical ability of an AI to understand or learn any intellectual task a person has, while superintelligence is defined by Oxford Professor Nick Bostrom as “any intellect that far exceeds human cognitive performance in virtually any area of interest”.
However, some argue that we are no closer to AGI, or superintelligence, than we were at the beginning of the century.
“You can say, and some do, ‘oh, it’s just around the corner.’ But the premise of that doesn’t seem very well worded. It was just around the corner 10 years ago and it didn’t come, “said Acemoglu.