Will it be AI vs the people? Or AI for all? Companies must build the social license they'll need.

Governments can't, shouldn't and won't force AI on an unwilling citizenry. Here's a short list of the big issues that are spurring public hesitation today.

Seattle just became the largest and the latest of hundreds American towns and cites that have moved to limit data centres.

Seattle is second only to Silicon Valley in experiencing the economic upside of technology, as home to both Microsoft and Amazon, 2 of the world’s 14 trillion dollar companies. Both companies sit at the centre of the boom in investment of Artificial Intelligence.

AI needs data centres. Lots of them. But as the race to build them is gathering momentum, so too is opposition. A Washington Post poll found voter "comfort with a new data centre in your community" was cut in half from 69% to 35% in just 3 years. Unusually, Trump voters and Democrats are mostly on the same page expressing discomfort.

Where will the water come from, where will the massive energy come from, and how much will the answer to these questions affect consumer costs. For those living near data centres, noise is another major challenge.

Hesitation about AI is related, obviously, but different. People worry about human job loss, and about the potential harm to humanity.

Last week the Carney government unveiled Canada's National AI Strategy, called 'AI for All.' Minister of Artificial Intelligence Evan Solomon — the former journalist and broadcaster who now holds the portfolio — has framed the strategy around building public trust.

Solomon’s announcement is an important step, the conversation needs shape, and voters need to engage on the issues, and contemplate the different and complex choices facing governments.

On the one hand, saying no to AI would be putting handcuffs on our economy. Saying yes - with no guardrails - would be just as reckless a choice.

Public concerns will not be easy to tamp down. Media coverage will tend to lean in on stories that document everything that can and might go wrong. People are naturally interested in things that could negatively affect them, and it’s hard for many to imagine the myriad ways in which AI could be beneficial.

The leaders of the AI enterprises often reinforce public anxieties. Anthropic’s CEO has said the world needs a brake, not only a gas pedal, when it comes to AI, because of risks to humanity that governments must regulate. Good for him to flag issues and point towards solutions. At the same time, this conversation will, for the foreseeable future, make social license for AI and data centres elusive and fragile.

The broad outlines of what it will take to establish a durable social license are clear. Here are the biggest question marks.

Water. Data centres consume massive amounts of water. Microsoft CEO Satya Nadella recently described how his company is working to solve this issue, creating a closed loop which would result in a data centre consuming roughly as much water as a restaurant. That would be a massively positive development.

Energy. Mind boggling is the best way to describe the tremendous amount of electricity data centres require. Where will it come from, how will it be paid for, who will pay for it, and will it be possible to use that much energy and reduce greenhouse gas emissions? All of these questions remain only partially answered, but what won’t work is the public paying more for their electricity in order to subsidise energy for hyper-scalers.

Noise. Massive data centres hum, continuously. They don’t make great next door neighbours. Finding the space to locate them where the noise pollution isn’t intolerable, or solving the noise some other way, will matter when it comes to community willingness to embrace a data centre.

Those are the main questions that surround data centres, and they sit alongside big questions about Artificial Intelligence.

What will happen to jobs? How can people be given the tools and support to adapt and use the technology quickly enough to keep pace with the displacement that will happen in some areas.

Is there any way to ensure ethical uses and avoid abuses? This is as much a question of technology as of law. Companies must be ready and willing to come up with innovative answers to this dilemma.

Can countries find ways to connect with the world while protecting their ability to act as sovereign states, on behalf of their citizens. In many parts of the world, America under Trump 2.0 feels threatening and untrustworthy, leading countries to look for solutions which limit US influence.

In my lifetime, alongside the invention of the Internet, there has never been a more important technological shift, prompting a more complex set of policy choices for governments.

What makes all of this even more challenging, is that the US under Donald Trump has abandoned the conversation about setting global standards and norms. MAGA America is racing to dismantle clean energy and deregulate technology giants. It’s like bringing a casino mindset to a brain surgery theatre.

It would be excellent for Canadians if our major political parties found ways to collaborate on solutions to these challenges, rather than to use AI as the next polarising wedge issue. Solomon’s approach has avoided partisan edges, to his credit.

Our country can be both a dynamic participant in the innovation and economic upsides of AI while thinking hard and being creative about protecting the public interest.

But in the end, government can’t do this on its own. Just as mining companies, banks, telecommunications companies, railways and many other sectors of our economy have had to develop, discuss, promote, and maintain visible standards that meet the public where they are, this will need to happen when it comes to AI and data centres too.

While the speed of capital formation for AI seems breathtaking right now, the effort to build public support, by the companies that will need it, feels like it remains mostly on the drawing board.