2021 AI Index: Four AI Trends That Stood Out

Visual China Group via Getty Images

March saw the arrival of the third edition of the AI Index, a comprehensive and well-respected report on the state of AI led by Stanford University and an interdisciplinary team from academia, industry and government. The report’s mission is to “ground the conversation about AI in data.” The most surprising thing about the report’s key takeaways was how unsurprising they were. Despite the epochal disruption of 2020 and AI’s accelerated adoption, the story of AI simply added another chapter and trends that we were already tracking continued to be important. As I looked through the key takeaways, four trends stood out to me:

1.     The center of gravity appears to be shifting to China, but that could be a mirage. Years ago, the story was that the US was falling behind in math and science. Today, we’re hearing similar stories about China’s successes in AI. The report shows that China has overtaken the U.S. in the number of AI journal citations, and .  However, the report also shows that the U.S. is attracting Ph.D. students and they’re staying in the U.S. after they complete their studies. It also shows that 65% of graduating Ph.D. students in North America move into industry careers. These two trends suggest that while China may be ahead in AI research, the U.S. will continue to pull AI talent from around the world (including China) and graduate them into the private sector, which will drive increased hiring and attract more talent. China may be taking the lead, but I’m bullish on the U.S.’s ability to win the marathon.

2.     New ways of addressing old problems will have far reaching consequences. The report highlighted drug discovery, content generation and surveillance as growth opportunities for AI. These are historically human domains that are increasingly being driven by artificial intelligence. The impact that these investments will have on our way of life, for good or ill, can’t be underestimated. Take drug development. A pharma company’s return on investment for research and development is at an all-time low ().  The promise of AI is that it can be used to improve this ROI by quickening the pace of discovery, and patients around the world will feel the impact of new therapies.

3.     Despite all the talk of ethics, we’re not seeing enough action. Two issues in the report highlighted areas where AI is struggling: Its lack of diversity and ethical consensus and benchmarks. We all know that diversity is sorely needed. Racial and myriad other identities need representation in AI. To build these teams, we must address both diversity and equity. Not only is it the right thing to do, but . Bias is not the only ethical concern that we’re waking up to. AI’s societal implications are profound and fortunately many are asking what can and can’t be done to protect humankind as we drive these advancements. We can’t wring our hands and do nothing. We also can’t woke wash and window dress our way through technological evolution. We must find consensus in the AI community, set benchmarks and do the work that needs to be done. This brings me to my final observation:

4.     Government regulation is coming, but we must not wait for it. As the report indicates, Congress has finally taken notice of AI. But it’s still Congress. Regulation may be years away and may arrive watered down or too restrictive. It’s up to the private sector to take steps to self-regulate immediately. They have the resources and the incentive. Sales will only suffer from the disruption that can be caused by unethical technology, and if they self-regulate adequately, the private sector may be able to avoid over-regulation.

If there’s any one thread that runs through each of these observations, it’s that AI continues to be a disruptive, world-changing force that will do both good and harm. It will influence geopolitics and transform societies. As the future’s picture becomes clearer, we need to prepare. We can’t treat this coming transformation with denial and hyper individualism. We’re in this together. A brave new world, with all its boons and dangers, is coming whether we like it or not.

Related posts

The Self-Optimizing Plant Is Within Reach


Business Contracts Remain Complex, And Artificial Intelligence Can Improve Contract Management


The AI-Powered Robot That Learnt Curling Using Adaptive Deep Reinforcement Learning