Technology (AI, blockchain, etc.) Archives

May 14, 2018

Using AI to Help the World Thrive

(I've gotten behind on re-posting my articles! I'll try to catch up in the next couple of weeks. As part of my new column in MIT Sloan Management Review, I've been diving into the world of new technologies and how they enhance or interact with the sustainability agenda. This is an article about AI, with a focus on Microsoft's efforts. More tech to come soon...)


What is the purpose of artificial intelligence?

The hype about AI, with its massive potential to disrupt business and society, is likely true. AI could make business radically more efficient and answer questions we didn’t even know we had. Of course, it may also destroy millions of jobs as machines get better than humans at everything from driving trucks to analyzing CT scans.

But focusing for the moment on the upside, it’s worth asking: Could AI help humanity solve its biggest problems?

Consider the challenges in front of humankind. We need to build a thriving economy and world for what the United Nations predicts will be 9.8 billion people by 2050. And we must do it without overwhelming our natural resources or making our climate uninhabitable. We’ll need dramatic changes in how the world works — deep shifts in energy, transportation, buildings, manufacturing, food and agriculture, and much more. We need to answer questions such as:

  • What’s the best, most economic path to a low-carbon economy?
  • How do we feed 9 billion or 10 billion people on a planet with a fixed amount of arable land?
  • How can we best move billions of people around crowded cities to keep those cities functioning, while using the least fuel possible?
  • How do we manage an electric grid with huge amounts of intermittent renewable power and billions of smart devices and electric vehicles plugged in?
  • How can our economic and political systems enhance well-being for all and reduce inequality?

We may need some serious help answering these kinds of questions. It’s quite possible that we’ve created complex, systemic problems that exceed our human capacity to solve them. In other words, AI may not just be nice to have; we may need it.

Some companies, particularly the tech giants, are recognizing this reality. They’re looking to AI as a tool for solving environmental and social problems.

For example, Google asked its DeepMind AI to examine the “complex, nonlinear” problem of how it uses energy in Google’s data centers (and this is no small issue: just in the United States, the tech sector’s data centers use 70 billion kilowatt hours of electricity at a cost of $7 billion per year). Google’s AI was able to slash energy use for cooling by an impressive 40%, saving significant money and carbon emissions.

In 2014, IBM launched a 10-year, $100 million project to use its Watson cognitive computing system to help Africa solve business and social challenges. The company is also leveraging AI to forecast solar and wind availability for power generation.

Enter Microsoft’s $50 Million “AI for Earth” Program

But perhaps most intriguing is the initiative that Microsoft recently launched — its own big play for leadership in the realm of “using AI to save the world.” In December, the company announced an expanded commitment of $50 million to, as Microsoft President Brad Smith wrote in a blog post, “put artificial intelligence technology in the hands of individuals and organizations who are working to protect our planet.” Smith pointed out that humanity is collecting a vast amount of data on the state of the planet. We need help, he wrote, to “convert it into actionable intelligence.”

The program, dubbed AI for Earth, is finding and funding innovators who are making progress in four critical areas — climate change, water, agriculture, and biodiversity. Microsoft’s first grantees, 35 teams from around the world, are impressive. The AI pioneers include a group in Italy using images of snow in mountains to better predict snow melt and thus water availability; the Jane Goodall Institute, which is helping “identify chimpanzee habitat connectivity and conservation priorities in Africa”; teams at Yale and Cornell using AI and data to understand crop health and improve yields; and a crowdsourced program, iNaturalist, that combines both “citizen-scientist” data with trained scientist input on biodiversity.

Microsoft will accelerate progress by providing seed money, intellectual support (in the form of a multifunctional team of AI and sustainability experts), and technology aid through its cloud computing resources. The AI for Earth program will also identify the initiatives that have the most promise and offer even more aid.

But the goal is more than creating some isolated success stories — it’s about being a catalyst for greater change. The stated mission of Microsoft’s AI efforts is “to empower every person and organization to thrive in a resource-constrained world.”

Rob Bernard, Microsoft’s chief environmental strategist, tells me that with AI for Earth, “we want to light up the ecosystem — we want the market to explode.” He imagines that once a team has created tools for, say, developing high-resolution maps of farmland from satellite imagery, other teams can build on it. They might ask different questions than the initial group, focusing on a different crop. Or look at a completely different problem outside of agriculture that could benefit from the same AI approach.

It’s a great idea. But a critical component of this “explosion of ideas” plan is making some capabilities part of a publicly available platform. So I have to wonder, what’s in it for Microsoft?

Business Payoffs for Being a Leader in Solving the World’s Problems

I see a few primary business benefits.

First, the initiative may help Microsoft attract and retain the best people. The competition for AI talent is intense and the tech giants are paying big bucks. Bernard says that when Microsoft posted some AI for Earth positions, some of the company’s top AI people jumped at the opportunity. There’s a clear trend, especially among millennials, for people to want more purpose in their jobs. Working on big, global environmental challenges is meaningful.

Second, the company can drive revenues for its cloud services. Digitizing the world, which we seem committed to doing, will require lots of data, servers, and software. Putting Microsoft in the middle of that whirlwind is good for business.

Third, the company could yield some related, but harder to measure, intangible benefits. Working on big issues and connecting to cool startups raises the company’s profile and keeps the 40-plus-year-old brand (I know, hard to believe) relevant and modern.

So, this whole movement will be good for humanity and benefit Microsoft (and other tech companies). And that’s more than OK. In fact, it’s critical to the success of the program. We need a large flow of ideas, capital, and talent to solve the world’s biggest challenges. Making it profitable to use AI in the service of humanity will attract more resources to the cause. Again, it’s likely that we need AI. Let’s just hope AI continues to need us.

(This post first appeared in MIT Sloan Management Review here.)

If you enjoyed this article, please sign up for Andrew Winston's RSS feed, or by email. Follow Andrew on Twitter @AndrewWinston

Andrew's book, The Big Pivot, was named a Best Business Book of the Year by Strategy+Business Magazine! Get your copy here. See also Andrew's TED talk on The Big Pivot.

May 30, 2018

A Blockchain Crash Course: How It Can Enable the Clean, Smart Grid

(Another article from my column for MIT Sloan Management Review. This is my attempt at an "explainer" on blockchain, describing it as simply as I can, and exploring how it may help build the clean grid. I will dive into other uses later...)


Blockchain is everywhere — at least when you’re talking about media coverage. It’s the new technology that enables those wild cryptocurrencies. But some techies think it will revolutionize more than money. In short, in the words of a recent Wired overview of the technology, blockchain creates “tamperproof databases.” Every transaction in some defined system, like all trades of a particular currency, is “cryptographically signed,” meaning it’s encoded to be theoretically unhackable. Then, the “block” of data, with its crypto info, is saved on thousands of computers in parallel — so that if you wanted to tamper with a record, you’d have to change everyone’s computer at the same time.

At first blush, it’s unclear why a really good, tamperproof database will be so groundbreaking. But big companies and startups alike are exploring how blockchain could change everything from contracts, corporate compliance, and auditing to digital identity and voting, supply chain tracking, and even building a smart, carbon-free electric grid. These ideas are fascinating, but the latter few in particular have caught my eye. Anything that might solve big, global, environmental, and social challenges, from climate change to reducing forced labor in the supply chain, is worth a look.

It’s impossible to unpack all of the potential ideas at once, so let’s talk about how blockchain might help with one critical aspect of the climate change fight: the challenges of managing clean, zero-carbon energy on the grid.

The pace of adoption of clean energy is breathtaking, but it’s not moving as fast as it could. One issue is that, as critics love to point out, the wind and sun are “intermittent” (they come and go). In parallel, how much energy we use swings wildly during the day. Demand in most regions peaks after dark, as everyone gets home from work, right when solar power is waning. This mismatch creates what the utilities call the “duck curve” problem, named for the misshapen curves of supply and demand. It’s a chaotic situation for utilities that are trying to keep a giant power grid in balance.

Massive battery storage will help solve the problem but getting to total grid scale is some ways off. We’ll also need demand-side options, such as a long-standing practice called “demand response.” When the utilities see fast-rising demand, they pay companies to reduce their energy demand by, for example, turning down their cooling systems a degree. Now imagine if billions of connected devices could adjust themselves to balance the grid. And then picture them talking directly to each other, partly through a shared, trusted system.

This is where blockchain may come in.

1. Blockchain Could Make Tracking Energy More Granular, Automated, and Trusted

To understand how this new tech could be a solution to thorny grid problems, I spoke to Jesse Morris from the Rocky Mountain Institute (RMI), based in Colorado. The nonprofit has cofounded, with the Austria-based startup Grid Singularity, the Energy Web Foundation (EWF), headquartered in Zug, Switzerland. The new organization is hoping to provide an open-source blockchain platform to the energy sector and accelerate the transition to renewables.

Imagine, Morris said, if every utility meter “assigned a crypto identity and time stamp, with geolocation,” to every unit of energy used and produced. On the supply side, the system would track every kilowatt hour of renewable energy generated. That would be useful to all those multinationals buying gigawatts of renewable power every year.

Right now, when utilities or companies buy renewables, they basically have three choices: Finance their own projects and own the panels or turbines, sign power purchase agreements (PPAs) to buy clean power from someone else for the next 15 to 20 years, or purchase certificates that represent the green attributes of energy. Some critics of these renewable energy certificates (RECs) question whether they truly drive new investment in clean energy — and thus whether the cleanly sourced electrons are actually displacing dirtier energy on the grid.

Morris believes that blockchain can help by making tracking more granular, automated, and trusted, which can allow companies to better verify claims of carbon neutrality. It will also help avoid double-counting on RECs or PPAs (that is, when an energy developer and the buyer of RECs count the clean energy as their own). So we’ll know how much progress on carbon reduction we’re really making.

2. Financing New Energy Projects Could Be Streamlined

Blockchain may also help increase the supply of renewable energy in another way. It could streamline the financing and insuring of new energy projects, which can get complicated and expensive. One new company I spoke with San Francisco-based, Banyan Infrastructure Corp., is piloting a blockchain ledger to manage the contractual obligations — of the developer, the buyer of the energy, the lender, and so on — and to lower some of these administrative costs. Since these costs are mostly fixed, smaller projects like a 10-megawatt solar farm would benefit the most.

But the real potential, according to Banyan’s CEO Tad Neeley, is to use blockchain ledgers to reduce the capital requirements for new projects.

The problem is this: Banks place conservative debt service requirements on the developers of renewables projects (because of uncertainties, some real and some exaggerated, about how dependable solar or wind farms are). The developers must keep a lot of cash on hand to ensure that they can meet loan payments. Neeley believes that a shared, real-time database that is tracking energy production can help lower the perceived risk and thus reduce these capital requirements. A shared ledger could make investors more comfortable that the turbines and panels will continue spinning off both electrons and cash.

3. Tamperproof Ledgers Could Facilitate a New Energy Market

So blockchain may help increase the supply of clean energy — but now consider the demand side. Millions of individual devices and building systems could track their needs and trade electricity device to device, across the full grid, or on small, localized microgrids. Blockchain’s shared, tamperproof ledger could verify all transactions, creating a new kind of energy market.

Of course, these visions of the near future create as many questions as they answer, such as: What’s the role of the utility when my refrigerator is talking to your HVAC system, or your solar panels are selling power to my electric car? This is not some completely futuristic scenario — the first blockchain-enabled sale of power from one neighbor to another just happened in England. And what, if anything, replaces the oversight and management of the shared resource that we call “the grid”?

But even with so much unanswered, bringing some excitement to this thorny problem of building a clean grid is a good thing in and of itself. It would be great to move beyond cryptocurrencies and bring blockchain innovation to cool, real-world problems like how every single thing that uses or produces electricity will communicate and optimize energy use. Or how to build a zero-carbon grid and tackle climate change.

This is all new territory, and I’m planning to dive into different parts of this story over time. The prospect of better tracking of supply chains, or of tracking carbon footprints for every product, is exciting. Stay tuned, as I’m sure there are newer blockchain applications — both real and hyped — to come.

(MIT Sloan Management Review first published this piece here)

If you enjoyed this article, please sign up for Andrew Winston's RSS feed, or by email. Follow Andrew on Twitter @AndrewWinston

Andrew's book, The Big Pivot, was named a Best Business Book of the Year by Strategy+Business Magazine! Get your copy here. See also Andrew's TED talk on The Big Pivot.