Happy New Year to all of you Eye on A.I. readers! 

If 2019 was any barometer, this year will be a busy one in the artificial intelligence world. The challenge will be to turn the advancements into something that businesses can use.

Most companies are still struggling to harness A.I. technologies like neural networks, or software that learns from data, according to surveys that Fortune covered last year. I believe this dilemma will continue in 2020, due in part to the lack of available A.I. talent that can build and maintain these complicated systems.

Despite the potential A.I. setbacks, businesses will continue spending big money on A.I. in the hope that their investments will eventually pay off. There’s just too much at stake (and too much hype) to ignore it.

Here’s my predictions for what to expect in A.I. in 2020:

1. The regulators are coming

For several years, lawmakers have been talking about regulating A.I.—but it will finally happen in 2020. For instance, on Tuesday, the White House introduced guidelines that federal agencies like the Federal Drug Administration and Department of Transportation must adhere to when crafting A.I. regulations, like whether their rules ensure that A.I. is “fair” to everyone. The principles, while vague, are a precursor to separate and more formal regulations that could impact how businesses use specific A.I. technologies like facial recognition.    

Even if federal A.I. rules fail to be enacted in 2020, local and state governments will continue to fill in the regulatory gaps, as cities like San Francisco have done for the use of facial recognition by law enforcement.  

2. A.I. chips go mainstream

After years of internal development and trials, you can expect startups to debut specialized computer chips for A.I. tasks like data training. Companies are increasingly seeking A.I. chips that consume less energy and that can crunch data more quickly. Also, startups may begin to challenge Nvidia, the leading maker of computer chips used to train A.I. systems.

3. Cashierless stores will debut worldwide

Fearing an increasing number of Amazon Go cashierless stores, retailers will debut their own versions. But they’ll likely lose money on these in the short term because the technology required to operate them is expensive.

4. China steps up in A.I.

China will play a bigger role in A.I. research in 2020. As Stanford University’s AI Index for 2019 shows, A.I. research papers from China are increasingly cited in other academic papers, underscoring their quality and importance. Jack Clark, an AI Index author and policy director of A.I. research group OpenAI, said Monday during an online presentation about the Stanford report that lawmakers believe that “the Chinese can’t invent things, they just steal things.”

“I sit there as someone who reads these research papers, and no, that’s very wrong,” Clark said. 

Jonathan Vanian 
@JonathanVanian
jonathan.vanian@fortune.com

A.I. IN THE NEWS

Are A.I. regulations coming? Over the next 18 months, the U.S. and Europe may introduce regulations impacting how businesses use A.I., the Wall Street Journal reported. The article included a statement by European Commission President Ursula von der Leyen, who said, “In my first 100 days in office, I will put forward legislation for a coordinated European approach on the human and ethical implications of artificial intelligence.”

Don’t send this A.I. software overseas. The U.S. government introduced a new rule intended to prevent U.S. companies from exporting certain kinds of A.I. software that could benefit “adversaries like China,” Reuters reported. The report said that companies must apply for a license to export certain geospatial imagery software from the United States to other countries except Canada.

A.I. hype hits healthcare. Kaiser Health News examined the burgeoning use of A.I. in healthcare and found that “many health industry experts fear AI-based products won’t be able to match the hype.” Among some of the problems the articles explores includes the fact that many health-tech startups do not publish their research in peer-reviewed journals, which means outside experts are unable to validate company claims.

A sad pizza party. Zume, a startup specializing in robots that make pizza, could lay off 400 employees as it aims to raise more money, Business Insider reported, citing unnamed sources. The startup has faced a number of setbacks and has attempted to move into new areas like "compostable packaging and food-truck services," the report said.

EYE ON A.I. TALENT

Groupon hired John Higginson to become CTO of the online deals and coupons company. Higginson was previously the CTO of the financial technology company Enova.

The Iowa Department of Transportation picked Matthew Rensch to be the state’s first chief data officer. Rensch previously worked at companies like Athene USA and Berkley Technology Services, according to the trade publication Government Technology.

EYE ON A.I. RESEARCH

A.I. that screens for breast cancer. Researchers from various Google units like DeepMind published a paper in Nature detailing an A.I. system that can help doctors more easily identify breast cancer by reading mammograms. In a New York Times article about the Nature paper, an outside expert developing similar A.I. technology said Google’s paper was strong, but cautioned that “the patients studied might not be a true reflection of the general population.”

FORTUNE ON A.I.

Holiday Superstar Wyze Suffers Data Leak, and Amazon Makes Returns Easy—By Don Reisinger

Rwanda Is Bringing Tech Buzz to Africa—By Richard Morgan

Why Workplace A.I. Needs a Human Touch—By Gwen Moran

BRAIN FOOD

Where’s the U.S.? Wired examines why the United States is not participating in an international consortium called the Global Partnership on AI, whose representatives from counties like Canada and France intend to “study and steer the effects of artificial intelligence on the world’s people and economies.” All of the so-called G7 countries have pledged support to the international group except for the U.S. “The White House says the body is unnecessary bureaucracy that threatens to dampen AI development by being overly cautious,” the article said.