What critics get wrong about the ‘American AI Initiative’
Share on Twitter
Sujai Hajela is co-founder, president and CEO of Mist, which develops self-learning wireless networks using artificial intelligence.
There’s been a bit of hysteria — AI-steria, if you will — over the Trump administration’s recently issued American AI Initiative, formally known as “Executive Order on Maintaining American Leadership in Artificial Intelligence.”
The initiative is a broad strategy “to sustain and enhance the scientific, technological, and economic leadership position of the United States in AI R&D and deployment.”
But critics have complained it’s short on specific actions and lacks new funding to accomplish its goals, in contrast to China’s 2017 “Next Generation Artificial Intelligence Development Plan,” which allocated billions to establish China as the “premier global AI innovation center” by 2030.
As Babson College professor Thomas Davenport noted in a recent essay in The Conversation:
One Chinese state alone has said it will devote $5 billion to developing AI technologies and businesses. The city of Beijing has committed $2 billion to developing an AI-focused industrial park. A major port, Tianjin, plans to invest $16 billion in its local AI industry.
“China is ready to leapfrog the USA in AI. The US initiative has NO money,” tweeted one skeptic, Moor Insights & Strategy analyst Karl Freund.
China is going to leapfrog the USA in AI. The @realDonaldTrump initiative has NO money. While US Startups are all finishing their chips, Huawei, Baidu, Cambricon, and Horizon Robotics are all shipping theirs. I guess the USA prefers to invest in walls.https://t.co/khJe26ndnj
— Karl Freund (@karlfreund) February 15, 2019
While I wholeheartedly agree that the United States must not cede AI leadership to China or any other country, I find some of the critiques of the initiative overwrought. In fact, even if its value is largely symbolic — more of a vision statement than a detailed blueprint — I still believe the initiative can help move forward the national AI agenda.
Let’s unpack a few things.
While many talk about AI in the context of a rivalry between two superpowers — and it well may be that, in some ways — AI is unlike similar global competitions of the past. Take the 20th century Space Race between the United States and the Soviet Union, for example. The effort never would have gotten off the ground, literally, without huge financial commitments from the U.S. government.
Money for AI’s advancement, however, is pouring in from the private sector. According to a report by CB Insights and PwC, venture capital funding of AI companies skyrocketed 72 percent last year, to $9.3 billion. The surge followed three years of steadily increasing investment, with a 28 percent average annual increase between 2015 and 2017.
And it’s not as though the government isn’t doing its part. IDC estimated the federal investment in cognitive and AI technologies is growing at a CAGR of 54.3 percent between 2018 and 2021.
Furthermore, while stopping short of specific dollar amounts, the initiative wasn’t exactly silent on the requirement for more government funding, calling on all relevant agencies to consider AI a top R&D priority and take that into account when developing budget proposals for fiscal 2020 and beyond.
Meanwhile, colleges and universities working on exciting research in AI and its enabling technology, machine learning and more are offering AI-specific training to help solve a skills shortage in the field.
College students reportedly enrolled in introductory AI and machine learning classes in record numbers last year, the number of academic papers on the topic soared and, according to a Stanford University analysis of transcripts, officials mentioned the technology in more than 70 congressional hearings.
I’m not surprised that Carnegie Mellon (no AI slouch, having introduced an undergraduate AI program last year) reacted very positively to the American AI Initiative. “The American AI Initiative’s focus on prioritizing research and development, responsibly leveraging data as a national resource and investing in an AI-ready workforce will bring new energy to our national innovation ecosystem,” the university said in a statement.
All of this points to the reality that the private sector, as has been the case so many times throughout the annals of U.S. innovation, is taking the lead in AI and counting on entrepreneurial spirit rather than government largesse to win the day.
As a Bloomberg editorial put it: “This contrasts favorably with (say) China, where the government is pumping billions of dollars directly into AI-related companies. This may advance the field somewhat, but it’s also a good way to sustain hopeless businesses, crowd out private investment, encourage cronyism, inflate bubbles, and generally make a hash of things.”
Simply by virtue of shining a bright spotlight on AI as a national priority, the initiative can have significant practical effects. For example, let’s say a VC firm is deciding whether to fund an AI startup or one in another hot space, such as the Internet of Things. The tone set by the initiative could tilt the decision in the AI firm’s favor.
It also could spur more universities to investigate interesting AI technologies in their labs, and further invigorate efforts to train the next generation of AI practitioners.
The American AI Initiative isn’t as cut and dried as the critics suggest. If it’s nothing more than a stake in the ground about AI’s essential role in the nation’s future, it’s still an important stake.