There’s no question that there are massive opportunities in AI for startups and investors. However, I think the real question is: Are there too many opportunities? Below I’ll cover where I believe those opportunities will come from.
Investment Opportunities in AI According to Experts
Recently, there was an interesting article on the Crunchbase website in which author Dimitry Smirnoff looked at investment opportunities in AI. Here are some points he makes:
- Just building an AI interface to a system or product is not enough.
- A startup can only succeed if it does something so spectacular that it gets acquired by one of the “giant“ firms.
- Companies must have a solid core business where they use AI to enhance their efficiency.
- The best values are in companies that facilitate AI.
My Experience with the Applications for AI
According to Investopedia, AI is a software heuristics simulation of human intelligence.
I’m not sure that the essential nature of AI has changed much since I did my PhD thesis work at Carnegie Mellon in the late 1960s. My advisors included some of the early AI pioneers like Alan Newell, Herb Simon, and Gordon Bell. I wrote my Ph.D. thesis on computer programs in Information Processing Language (5). This was a program created by Newell, Cliff Shaw, and Simon at Rand Corporation and Carnegie Institute of Technology in the 1950s.
They designed it to simplify the processing of lists of information, which is the essence of AI.
Of course, I’m sure that designers have developed new heuristics. But the major advance that has enabled the current state of, and brings more opportunities in AI is the enormous increase in computer power. Moore’s Law, the creation of Gordon Moore, one of the early employees at Intel, says that the amount of circuitry that engineers can place on a computer chip doubles approximately every two years.
It does not translate exactly to power increases, but it may not be far off. Therefore, if we assume that computer power doubles every two years, then today’s computers would be approximately 270 million times more powerful than those of 1968. To state it differently, one of today’s computers can do a task that would have required over 3000 days in 1968 in about one second.
New Opportunities in AI Arise with More Computing Power
With these advancements in computing power come new applications and new opportunities in AI.
For example, consider the advancements in chess-playing and natural language voice recognition programs. In the 1960s, Alan Newell’s championship chess playing program required hours to look ahead five or six chess moves. Today’s programs can look ahead by fifteen or twenty moves.
Even in the 1980s, the voice recognition software company I invested in could not operate in real-time. Yet today’s cell phones have voice recognition technology that translates as fast as we can speak.
This power, combined with advances in heuristics, opens up enormous commercial opportunities in AI. I see this every month in the startup pitch decks I receive, as many companies are beginning to use AI to implement their ideas.
There’s A Risk of Overinvestment and Oversaturation
It seems obvious that AI will be another area where venture capitalists grossly over-invest. This happened with disk drives, ASICs, internet applications, software, biotechnology, and phone-based apps. The venture capital industry has a history of over-investing in hot areas.
In my experience, top-tier venture capital funds often cherry-pick the earliest and best deals. This is partly because they have good eyes and because good projects get referred to them. The fact is, the top venture capital firms tend to cherry-pick the early winners and leave less competitive companies for the rest of the industry. Therefore, opportunities in AI for startups are limited without a first-mover advantage.
The wave of AI companies may be greater than any of the previous waves because the cost of entry is incredibly low. Companies using existing AI engines can assemble sophisticated systems with little effort. It’s almost as simple as asking an AI machine: “Create for me a system that does the following…”
Many investors may sit on the sidelines and wait for the shakeout. Moreover, there may also be some concern about bias in the AI engines and whether or not that would affect their commercial attractiveness.
Who Will Benefit from these New Opportunities in AI?
In the end, I think the winners will be companies with:
- First-mover advantage.
- Significant market traction.
- Enormous amounts of capital for capturing and maintaining market share.
Some venture capital firms will invest for fear of missing out on the opportunities in AI if they don’t. However, I think the winners will require very large sums of capital.
In conclusion, Dimitri Smirnoff probably has the right idea that many winners will be companies that provide the engines for AI rather than the applications.
Do you have an AI focused startup? Will you simply be applying AI in the structure of your company, or will you have a first-mover advantage? If you’re based in Southern California you can submit your pitch deck to me for critique via Monday Club. You can also learn more about what makes a startup fundable by investor’s standards in my book The Fundable Startup.
Recent Comments