Machine Learning Leaps Forward

The Terminator is an American film classic.

The movie’s villain is a military AI (artificial intelligence) program named Skynet.

The system learns at an exponential rate and eventually becomes conscious, all-powerful and dangerous.

I won’t spoil the end of the movie, but it doesn’t work out great for humans.

The AI we have today is nowhere close to Skynet.

What we have now isn’t really “AI” at all in a strict sense. Even IBM’s famous Watson doesn’t make the cut. Wikipedia calls it a “question-answering computer system.”

A simpler way to describe what we have now is “machine learning” (ML). Smart software programs that learn and adjust their behaviors accordingly.

This is a technology that all investors should be familiar with.

How It’s Used

Personalization is a major way ML is used. Customizing the customer experience in a way that could never have been done manually.

So when Netflix recommends certain films to you, that’s ML.

Siri, Google and Amazon too. Machine learning is a core part of all these businesses.

ML is also quite a hot feature in the startup space. Especially when combined with large (preferably proprietary) sets of data.

Why? Because well-executed ML can make a good business great. And it can make a great business an outstanding one.

Google Leads

I think it’s safe to say Google is the most advanced company in the world when it comes to ML. Email (spam), maps and navigation, and search. All these services are constantly learning how to deliver the most accurate and satisfying results.

For example, if enough people misspell a word when searching, Google will recognize the typo and deliver results based on what everyone’s really looking for.

Apple’s virtual assistant Siri is another example of machine learning at work. Siri uses natural language processing and other ML techniques to deliver the most accurate results.

ML is about automating improvement. Utilizing large data sets to accomplish big things, in essence.

ML Startups: An Area to Watch

In the last few years, machine learning has made amazing progress. New open-source tools have emerged that make it easier for developers to build their own ML projects.

Just this month, Google released its own internal tool for deploying ML algorithms. This led a number of other high-profile companies to open-source their tools as well.

In other words, it’s a great time to keep an eye out for startups using smart ML to accelerate their businesses and build better products. Used correctly, it can be an almost unfair advantage.

Machine learning can be useful in most businesses. It’s just a question of whether the company will grow large enough to justify the (significant) cost.

In recent months, I’ve had discussions with a number of startups that plan to extensively use ML.

One of them is a social app for meeting friends out at a bar or restaurant. The product will be making recommendations and suggestions to users based on their past actions. So strong ML is a must.

This is a great example of what I’d consider a “machine learning startup.” Key features of the product require strong ML talent. Luckily, one of the co-founders wrote code for a large hedge fund in New York.

ML is a specialized science, so you want to look for strong technical experience in the team.

The social app startup looks extremely promising, and I’ll tell you more about it once it’s out of beta-testing. I’m currently lobbying to convince the founders to raise money via equity crowdfunding.

And it’s going pretty well. It isn’t hard convincing a founder that having 10,000 new investors and $1 million is good. But… convincing someone to try something new and different can be.

I’ll keep you posted.

Good investing,

Adam Sharp
Founder, Early Investing

P.S. The SEC has released a bulletin announcing that Title III equity crowdfunding will go live May 16. Access to true ground-floor opportunities is now just a few months away.

We just published a brand-new comprehensive guide to startup investing. It includes all the information you need to know to get started. Where to invest, what to look for and which red flags to avoid. I highly recommend getting a copy for yourself, and I suspect it will pay for itself many times over. Click here to find out more.