The Keelwood Issue #3 - The AI-first country

Hey fellow followers. Still cliffhanging? Beans will now be spilled.

Binary Brains is changing its name to Appetite to better vibe with our focus on the restaurant industry. Keep an eye out for a new brand and webpage. That was it. Time to let go.

AI in Sweden

I’d also like to bring up the topic of AI in Sweden. I actually started writing this before the Microsoft announcement this week that they will invest in a data center in Sweden with 20,000 Nvidia GPUs (if I remember correctly). But the topic of AI became even more relevant this specific week. And make no mistake, this doesn't change that Sweden is lagging behind in regards to AI maturity.

As Marco Iansiti and Karim R. Lakhani put it in the book “Competing in the Age of AI” about the age they claim is the age of AI:

"For the first time in more than a hundred years, we are seeing the emergence of a new kind of firm, which we claim is defining a new economic age."

Or why not Gil Pratt, who works as Head of Toyota Research Institute on AI:

“The biggest thing since the Cambrian explosion 500 million years ago.”

Something big is happening, and we all have a responsibility to keep up. It is a fourth wave of AI, and it has gone mainstream. But while the Everyman doesn’t really understand what AI is, Everyman are working to adapt to technology they don’t really understand. It’s worst with the politicians, who don’t really understand anything about AI and at the same time have the greatest responsibility to prepare our country for this new technology. And as you might know, it’s not going very well.

I saw some ranking on how well countries have matured in AI. Sweden is in 17th place, if my memory serves me right. Instead of checking this ranking and ensuring I remember correctly, I started looking up a bunch of other rankings instead. Fittingly, we in Sweden are somewhat good at most things (called "lagom" in Swedish). We are, for example, #9 in entrepreneurship where we lag behind due to the difficulty of accessing capital and our lack of technical expertise (the Everyman and Everywoman ranked in tech understanding, including politicians).

Another example is "Adventure," where we are #24. This is because we are not "Sexy." If you have questions about that, I refer you to the source to file complaints. It is based on someone’s very objective definition of what is Sexy and an evidence-based correlation analysis on how Sexy relates to Adventure (or not—again, ask them). "Adventure" also loses out because of our lack of humor and the fact that the weather sucks. We can at least agree on that. We are #14 in "Cultural Influence." That’s because we are "Happy" but not "Trendy." Our culture also lacks "Impact." Someone should inform the ranking institute that Max Martin is Swedish.

When it comes to rankings that really matter, we are absolutely the best. We are #1 in Quality of Life and #1 in Social Purpose. The ranking that summarizes rankings places us at #3 in the world. It is the ranking of rankings. The rank that summarizes rankings that summarizes rankings.

So, it is strategically smart to be moderately good ("lagom") at everything (except Sexy, Fun, and Technical Expertise where we suck) to be the best at what really matters.

But while the politicians have done a regression analysis on these rankings and concluded that "Lagom" (read rank #17) is best when it comes to understanding and investing in AI to maintain our Social Purpose and Quality of Life-rankings, I would like to argue the opposite.

How about becoming the best in AI in a different way than everyone else?

While the USA and China are investing in having all the capabilities for AI—a match we will never win—we can instead become the best at understanding and applying AI to our actual operations. It is clear that within the EU, it is not possible to grow companies like Open AI, Google, Microsoft, or similar. Otherwise, we should have had something similar by now. And no matter how cold it is in northern Sweden, and how good our electricity production is there, we will still not be able to build data centers with more Nvidia GPUs than what the USA's or China's investments in the area have (Microsoft can, however—although it is not that many in comparison—but it is, of course, good to have the infrastructure here in our vicinity).

But what about raising our level of technical expertise? To start educating our children in Tech and AI? To start educating all medical students, economics students, social work students, teacher students (etc.) in basic AI knowledge with the possibility of deepening it? To start educating all management teams, boards, and all employees in Sweden’s fine successful and failure-rich companies to discover the opportunities with AI in every nook and cranny of our country. And to become the best through a thousand AI-cuts instead of huge investments. It will be the home computer of the 90s all over again. AI for all. To put AI to work at a level and frequency like no other country. Giving everyone education so that we understand what we are doing, how AI can be used, can't be used, should be used and shouldn't be used, and how we conduct accompanying change work.

That I believe in! Because then we will not only replace many jobs but also create many jobs. Or as Brynjolfsson and McAfee put it in Harvard Business Review:

“...[AI] will not replace people, but people who focus on AI will replace those who do not.”

And even if we get better at AI and tech, I don’t think we will lose our Social Purpose or Quality of Life-rankings. While politicians are busy arguing about how little money the Swedish school should get, I argue that the first jobs we replace with AI are all politicians so we can vote on different algorithms instead. DemocrAI! My algorithm you could vote on would only be trained on evidence-based research (which the above argument is not based on)—free from party-political points to increase the chances for the next election. Instead focused on optimizing our welfare and efficiently using our tax money.

Now you might think that it wouldn’t be entirely ethical. What happens if this and that. I would be naive to say that I don’t see risks with it. If nothing else, then I can at least educate the politicians instead, and we can develop AI models available for everyone to see what we need to do to optimize welfare and the future for Sweden. Just as I argue that all organizations should adopt an AI-first Company approach, I think we should become an AI-first Country.