Recently CB Insights has published its annual Report on the Top 100 Artificial Intelligence Startups redefining industries.

The full report can be downloaded here and contains a hand-picked selection out of nearly 5K startups, based on several factors including patent activity, business relations, investor profile, news sentiment analysis, market potential, competitive landscape, team strength, and tech novelty.

This is the short list of selected companies:

Top 100 AI Companies for 2020 According to CB Insights

Let’s dive deeper and learn what this data can tell us about the macro status of the AI sector. We’re not interested so much here about the specific phenomena (whether this or that company was selected) but on the underlying trends at a larger scale that the data points to.

Top AI Countries

If we break down where these companies are located, the hegemony is pretty clear: United States is home to 6 times the number of Top AI startups than its closest competitor, with Canada, UK, China, Israel and Germany as distant followers.

Click on the chart for interactivity

Of course, not all these companies were born in USA. Undeniably many of these startups relocated their headquarters there once they had enough early traction and funding to justify doing so. Many successful investors across Israel, UK and Northern Europe are capitalizing on that strategy. Precisely the fund I am involved with at the moment, Conexo Ventures, capitalizes on that strategy for the Southern European market.

But besides national pride, that point should be moot for entrepreneurs: the reality is that once a technology ecosystem is established with such an undeniable leadership, its gravitational force pulling in money, talent and experience is strong enough to provide a huge advantage to the local players and preclude any other clusters to solidify. Wise founders should understand this and get ready to re-domicile should the opportunity arise.

This is the stark geo-political situation, now plotted on a world map.

Click on the map for interactivity

Top AI Funding

If we pay attention not just at the raw count of startups but at the dollars fueling these companies (that is, aggregate AI funding per country) the picture does not change dramatically but reveals an interesting subtlety.

Canada, while being 2nd in quantity of startups, it drops to 6th position when ranked in terms of funding, right behind its competitors UK, China, Israel and even Germany, right before the long tail.

Pretty cool. In terms of efficiency, which is a very important factor in ultimately deciding the rentability of any investment, Canada deserves a shout out, well done guys!

Click on the chart for interactivity

AI in Spain

After the top six contenders there’s a longer tail of countries where only a startup made it to the list. Spain made it to the list thanks to Sherpa, which builds AI technology and products such as Digital Assistants, Recommenders, as well as more generic AI technology, such as Federated Learning.

Sherpa originated in Bilbao, home to its founder Xabi Uribe-Etxebarria, which is particularly pleasant for me since I was born in neighboring Portugalete, close to its renowned hanging bridge. Now we both have swapped bridges for the golden one in Silicon Valley and Xabi hails from Palo Alto. He nevertheless has it clear that the fight is not over, the stakes are just larger, and that he needs to persevere in the strategy that led Sherpa to its success:

“The key is in specialization. Our competitors have different sections inside the same company, hey build from search engines to autonomous cars to desktop computers. Really we are competing on a very specialized section, but one that has great impact inside these companies. We cannot compete with the entire company, but we can against those specific sections, like AI, and perform better than they do.”

Top AI Investors

Let’s switch gears and have a look at who are the leading AI investors worldwide. For that purpose we define “leading investor” as those with more companies that made it to the list — irrespective of the amount invested. Most likely these are going to be all USA funds… but which ones?

And the winner is… Google Ventures, with 8 investments. Following closely on their heels with 7 investments we find Peter Thiel’s own Founders Fund.

Click on the chart for interactivity

Then we have a tie up for 3rd place with 6 investments each among Khosla Ventures, Data Collective, Sequoia and Plug and Play, the top Silicon Valley accelerator with branches around the world (among them Plug and Play Spain located in Valencia). Precisely these days Plug and Play is celebrating its Spring Summit 2020 so I recommend that you stay tuned to see what is brewing in this unicorn powerhouse.

Top AI Industries

Finally let’s close up with by digging on the industry distribution. This is always a hairy topic as there are so many ways to classify them, including the usual confusion between Economic Sectors (which is our real interest) and Technologies (methodologies or techniques used) on the one hand, and Business Models (like selling to businesses or to consumers) on the other. I will spare you the clean-up details, but I recommend that you check the supporting notebook for details.

Click on the chart for interactivity

Unsurprisingly, the most popular industry by far is Software (the “mother of all industries” for AI). But is is somewhat surprising to see Manufacturing and Transportation ahead of other sectors where AI is mucho more spoken about such as Finance, Insurance and Real Estate. Indeed Manufacturing includes which includes Robotics and Industrial Automation, while Transportation includes Autonomous Vehicles plus Logistics and Navigation that are still capturing the lion’s share of Top AI Startup dollars.

There is an interesting correlation here, that goes from very organized environments (a software program, an assembly line, a route), to reasonably organized (Real Estate, Insurance, Finance) to the least organized or, inversely, those in which context variability requires human handling, such as Environment, Education, Agriculture and Human Resources.

It is likely that as methodologies continue advancing and algorithms are able to handle higher amounts of chaos, maturing closer to a “general AI” rather than “context-specific AI” we will see increased success in these areas that are now at the tail of the distribution.

Image by Gerd Altmann from Pixabay

Supporting Documentation

If you want to explore the supporting documentation and even play yourself with the data directly, check the following links: