Just a few years ago, smart cities, a catch-all phrase for connecting urban infrastructure like parking meters and stoplights online, had major buzz. The idea was to use machine learning to crunch data collected across cities to, for example, minimize traffic delays and reduce pollution.
But smart city projects haven’t taken off as quickly as expected, partly because of COVID-19 decimating local economies and reducing city budgets.
One of the latest smart-city causalities was Cisco’s ambitious Kinetic for Cities software, a dashboard for governments to manage data from their local smart-city projects. Cisco said in December that it would kill the product.
“The cities realized that that’s not what they actually need,” said Cisco global innovation officer Guy Diedrich.
Instead of pushing its software, Cisco plans to work with city governments on other efforts involving Internet connectivity. The company recently worked a lot with local officials to install basic Internet infrastructure, for instance, like helping Brazilian cities create virtual courtrooms, he said.
Forrester analyst and smart city expert Michele Pelino said local governments are feeling pressure to keep city operations humming despite shrinking budgets. Because smart city projects can be expensive and often don’t pay off for years, they’re being cut back.
Furthermore, some smart city projects had been focused on fixing problems that are no longer as pressing. Reducing traffic and parking now in many cities isn’t as important because many workers no longer commute to downtowns because of shelter-in-place rules.
Now, city governments see “security and safety” as a priority, and have reallocated some of their budgets to solve the problem, Pelino said.
“People need to feel comfortable that they are coming into an environment that is safe and secure,” she said about these kinds of projects.
Pelino is unsure when more ambitious projects, like using machine learning to optimize traffic lights, will be reinstated. She’s “certain” that bigger projects will eventually come back, but for now, many cities don’t have the money.
“Where is the money going to come from?” Pelino said.
As for Cisco, shuttering its smart city technology is another example of its challenges in expanding beyond its core networking gear business. Although the networking giant has been pushing into new areas like selling more specialized chips and networking software, the company’s overall business is still shrinking as firms buy less routers and switches than they used to, partly due to the rise of cloud computing.
Diedrich is still optimistic about Cisco’s smart-city push, saying that the company must still learn about it by working with governments via its Country Digital Acceleration (CDA) program, intended to help various governments with their digitization projects.
“What we learned with coming to cities, is that their needs are evolving and changing in real time,” Diedrich said. “One thing that we’re not doing is coming up with what we think cities need and shoving it down their throats—that’s not Cisco and certainly not the CDA program.”
As to whether Cisco would have kept selling its Kinetic for Cities software if it wasn’t for the coronavirus pandemic, Diedrich was vague. One thing that is certain: any smart city projects the company pursues will involve artificial intelligence.
Among the local governments still testing smart city projects, officials “know A.I. and machine learning is going to be very important,” said Diedrich, adding that it’s needed for making quick decisions like adjusting traffic lights quickly.
Jonathan Vanian
@JonathanVanian
jonathan.vanian@fortune.com
A.I. IN THE NEWS
Step on the gas. General Motors and its self-driving car startup Cruise are partnering with Microsoft, which is joining GM, Honda, and other investors in a financing round of over $2 billion. As part of the partnership, GM said that Microsoft would become its “preferred public cloud provider,” helping the company to “accelerate its digitization initiatives, including collaboration, storage, artificial intelligence and machine learning capabilities.” With the investment, Microsoft joins tech giants like Google parent Alphabet and Amazon that are spending billions of dollars on self-driving car projects and businesses.
Slap on the face. The Federal Trade Commission has ordered the company Everalbum to delete any photos the company secretly obtained of people’s faces as well as the machine learning algorithms that were developed using those photos, tech publication OneZero reported. The FTC slammed Everalbum in a statement saying that the company “made promises that users could choose not to have facial recognition technology applied to their images, and that users could delete the images and their account.” The FTC also cited how the company dealt with the tough data privacy and facial recognition laws in some states by instead focusing on people living in states without the rules. From the statement: Because the people of Illinois, Washington, and Texas passed laws related to facial recognition and biometric identifiers, Everalbum took greater care when it came to these individuals in these states. The company's deception targeted Americans who live in states with no specific state law protections.
Facial recognition for the masses. Citizens are using facial recognition software to analyze videos of the Capitol riots and identify people who took part in the havoc, Vice’s Motherboard tech publication reported. One of the technologists who the publication spoke to said they have shared their findings with the FBI. The report noted that facial recognition software “can often be inaccurate” and that citizen-sleuths using the software are an example of “the democratization of facial recognition.”
China on the A.I. rise. A new study by the Information Technology and Innovation Foundation thinktank about governments and their development and use of A.I. shows that the U.S. is still the leader in A.I., followed by China and the European Union, the South China Morning Post reported. Although the report said that the quality of U.S. A.I. research papers is better than China and the E.U., China is still investing heavily in A.I. and “has made strides in several areas and last year had more of the world’s 500 most powerful supercomputers than any other nation – 214, compared with 113 for the US and 91 for the EU.”
EYE ON A.I. TALENT
Oracle named Don Johnson as the database giant’s head of a new cloud and
A.I. unit, Business Insider reported. Johnson was previously executive vice president of the Oracle cloud infrastructure unit.
Bright Machines has picked Abhishek Pani to be the industrial automation technology startup’s chief product officer. Pani was previously the startup’s senior vice president of product management and was also a former Adobe senior director of A.I. and data science.
Handshake, a startup focusing on job recruitment for college students, hired Asif Makhani to be its chief technology officer. Makhani was previously a director and general manager for Google and its Google Images unit.
EYE ON A.I. RESEARCH
A.I. at light speed. Researchers from Monash University, Swinburne University of Technology, and RMIT University published a paper in Nature that explores the possible role that optical computing may play in deep learning. It’s a wild idea that’s based on “harnessing the distinctive properties of light” as a way to speed up the training of neural networks and other related A.I. tasks. Although this remains research, it does offer a glimpse into the possible future of specialized A.I. chips.
From Nature: Such a highly parallelized framework can potentially process an entire image in a single step and at high speed. Moreover, in principle, the system can be substantially scaled up using commercial manufacturing procedures and aid in situ machine learning in the near future. Because the convolution process involves passive transmission, the calculations of the photonic processing core can, in theory, be performed at the speed of light and with low power consumption. This ability would be extremely valuable for energy-intensive applications, such as cloud computing.
FORTUNE ON A.I.
Google takes action against another outspoken A.I. ethics researcher—By Jeremy Kahn
How two of India’s richest men became the target of farmer boycotts—By Naomi Xu Elegant
Which mobile carrier has the best 5G network? It depends—ByAaron Pressman
The technological hurdles of onboarding new employees in a pandemic—By Chris Morris
Facebook passes the buck on decision to ban Trump to its new Oversight Board—By Danielle Abril
BRAIN FOOD
Inside the rebellion. A little-known startup Rebellion Defense is the subject of an interesting profile in The American Prospect, which explores how the Eric Schmidt-backed company has become a player in Washington in a short amount of time. The article explores Schmidt’s interest in advocating for the Pentagon and other intelligence agencies “to adopt more machine-learning technology” and how Rebellion Defense plays into his strategy. The article quotes an unnamed venture capitalist who questioned the startup’s tech chops, saying, “If you can offer a big hammer with the letters ‘AI’ on the side, they’ll hand you a bag of money,” referring to the gullibility of some government agencies. Regardless, the startup appears to be gaining some momentum and is taking on A.I. military projects that employees from Schmidt’s former company Google refuse to work on.
From the article: Rebellion called themselves “a modern day Manhattan Project.” They emphasized their recent knowledge of the Pentagon and Congress as a “substantial head start to the early-stage startups attempting to build products that are to be sold into government.” Their three initial labs would include using AI for the military and policing, protecting large data systems, and creating the ultimate antivirus system. Each one would be worth billions of dollars. “The mission of national defense must be the place to be in tech,” Rebellion’s founders wrote. “This is an unconstrained ‘Project Maven.’” Rebellion would do what Google would not. The pitch seemed tailor-made for Eric Schmidt.
