Governments in the U.S., UK, and EU have parted the metaphorical AI waves for Big Tech – and these giants are walking through while the rest of us tread water in the ocean.
Governments are regulating – and deregulating – specifically for these firms, reorganising national energy infrastructure for them, and spending billions of taxpayers’ money on the massive infrastructure projects they’re asking for.
Under Trump alone, U.S. states have faced a 10-year ban on introducing AI regulation, and over a dozen federal sites have been designated for datacenter development.
Big Tech valuations are skyrocketing. Meanwhile, most businesses are failing to adopt AI or to get a return on their investments in it.
To me, it looks increasingly obvious that our governments have sold us and AI out to Big Tech. And it’s high time they switched the script and started building their AI strategies around supporting the everyday enterprise, startup, and SMB.
Governments have signed up to the Big Tech vision in part because it’s flashy. It’s cool. It makes headlines.
But it’s not just a publicity thing. Our leaders, utterly unable to grasp the technological foundations, let alone the economic and societal implications of AI, have looked around for experts who can give them a vision. And who better than the single-minded Big Tech execs.
These execs sell them one clear vision. We need to pull out all stops to build the largest and most powerful frontier models and datacenters, and everything will be solved in the process. We’ll ensure digital sovereignty and prosperity. We’ll fend off competition from China. We’ll even solve climate change, just before we’ve accelerated it beyond the point of no return.
They give our leaders a crib sheet of their demands, and our politicians think they can get away with not doing their homework. They haven’t looked into what it means to build a thriving AI ecosystem. They haven’t looked at the many different forms AI takes, and which types of AI are economically beneficial in different contexts and industries. They haven’t looked to solve the concerns and challenges faced by everyday businesses trying to successfully adopt AI.
They clearly don’t understand what’s needed to lead their countries through a successful digital transformation that benefits the entire business ecosystem and populace, rather than a small group of monopolistic companies.
The alternative? They could actually listen to and take action on behalf of AI execs at enterprises and startups, CEOs at SMBs, and the leaders of large non-tech enterprises. They could listen to economists, social scientists and AI ethicists.
But this group is large and diverse. They don’t have a single challenge, solution, or vision – but a naturally complex set of demands and concerns that need to be addressed with thoughtful policymaking. And this is apparently too much for our policymakers to handle.
Our leaders have opted for pre-made to-do lists over technical depth and understanding. Big Tech’s vision over the needs of the rest of the business ecosystem.
So, what’s the solution? How do we flip the script away from Big Tech’s AI vision, towards the needs of the rest of business?
First, politicians need to break away from viewing AI entirely through the lens of Big Tech’s narrative: that AI success equals the biggest and most powerful models and datacenters. Because AI is far bigger and more complex than that. Frontier models, like today’s media darling LLMs, are the jumbo jets of AI – and governments have forgotten the family cars, trains, and tankers that are integral to the economy too.
Once they’ve opened their eyes beyond LLMs, they need to work across the ecosystem to identify the barriers – talent, data access, funding, regulation – businesses are facing to building these AI technologies into their operations. Then, they need to make breaking down these barriers a higher priority than cow-towing to Big Tech’s demands.
Second, they need to open up their data to these businesses. There’s no AI without data. Businesses need high-quality, diverse data on the populations they’re serving. And making state-held data safely available as a production factor for business’ AI development is a strong step towards giving them what they need to build solutions to address real-world challenges.
Third, they need to incentivize Big Tech to open up its own data treasure troves – or utilize antitrust measures to do so. This will be far harder for governments in the UK and EU, which have no jurisdiction over these companies, but they can exert pressure when it comes to data involved in operations in their country.
This data gives Big Tech an uncompetitive advantage over any other business trying to train and develop AI, and this would go a small way towards levelling the playing field. Of course, everyday businesses will never outcompete Big Tech at scale with AI – but if they’re given the right building blocks, they can get tangible value out of it.
Finally, governments need to drop their obsession with frontier AI development, such as LLMs, and start thinking about the interior. Instead of constructing a ladder for Big Tech to keep chasing its AGI mania at taxpayer and climate expense, they need to build the scaffolding for AI to grow outward – so it can benefit every business, and every customer.
This means building a thriving and broad-based AI ecosystem – not a top-heavy skyscraper for tech giants to dictate from. They can do this by focusing on talent, building safe and effective data-sharing frameworks, and regulations to support competitiveness.
This might all sound unflashy. And rightly so. Cars, trains, and tankers are less interesting than jumbo jets, but far more important to everyday transport. We need to apply this rationale to AI.
It’s time our leaders stopped shaping their AI strategies for 1% of companies – and gave some thought to the other 99%.
The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.
