How to start an AI business

AI-based businesses are so enticing, to investors and customers alike, that a large number of companies have started to pretend their human employees are machine learning algorithms.

The rise in ‘pseudo-AI’ has seen businesses offering services that they claim are based on artificial intelligence, but which are really rooted in the actual intelligence of normal human beings. Edison Software, for example, sells an app that suggests “smart replies” when users are writing emails. Rather than using an algorithm to do this, the firm employs people to read private emails and type out replies manually.

The big reveal that companies like Edison don’t really use AI has been a PR disaster for all involved. Combined with the clear ethical violations, this is more than enough reason not to take the pseudo-AI route yourself.

Instead, those who want to strike while the AI iron is hot should follow these tips to starting a genuine AI business.

Solve a common problem

This is good advice for anyone starting a tech business, AI-focused or otherwise. But since AI offers so many more solutions, you can cast a wider net in terms of what kind of problems you are tackling.

Don’t just start an AI business merely for the money. You have to use it to tackle something you are passionate about, and that you think others will be too. Identify the problems potential customers might be facing and orient your business to solve them.

Spotify uses AI to introduce listeners to their new favourite bands. IBM uses AI to win Jeopardy. On the smaller end of the scale, some of the most exciting AI startups at the moment include Vivacy, which uses machine-learning cameras to improve road safety, and Lobster connects brands with relevant user-generated social media content for advertising purposes. There really is no limit to what you could do with AI, so think creatively about what products or services your business might provide.

Hire experienced staff

AI may be new, but it’s been around long enough for people to have gained substantial experience in the field. Unless you happen to be an AI coding guru yourself, it is essential to have people around you to develop new software and applications.

Finding the ideal candidates for your business will depend on what direction you are pursuing. Niche recruiters like Eursap, who specialise in SAP jobs, will be able to find you experts in SAP applications and development if you’re thinking of focusing on an AI-powered SAP solution. For more of an all-rounder, tech recruiters like TechnoJobs have boards and candidates dedicated to finding AI-savvy employees.

Act fast

Artificial intelligence is not a bubble that’s about to burst; it’s only likely to grow as time goes on. But that doesn’t mean you shouldn’t hurry. Now is the time for your business to get in on the ground floor. Though large companies like Google and Amazon have their own AI divisions, they’re likely to be paying large sums of money to purchase the startups who are really breaking new ground.

If your new business is successful enough, you might find yourself the subject of a bidding war between Apple and Google as dozens of companies have been over the past five years. These two tech giants in particular have been buying up promising AI firms from around the world, including UK startup DeepMind, which went on to become Google’s key machine learning division after being purchased in 2014.

Clearly, now is the moment to develop an AI business, so you should waste no time founding your startup, hiring the right staff, and priming it for financial success. With any luck, you’ll end up running the next DeepMind, and when Google comes calling, you can either cash out, or press on to build a tech empire of your own.