Startup’s Parable of Rocket Ship and Fuel

This is a Short Note.

It is indeed happening. Everyone wants to get into AI  (artificial intelligence) because that is the buzz of the time. AI is cool and fascinating. Across African cities, our entrepreneurs are pushing, just as their counterparts around the world.

Unfortunately, there is one big problem: you cannot just get into AI because you know maths, statistics and algorithms. Succeeding therein demands more. AI is the rocket ship. It needs a fuel. That fuel is data. Without data, AI is simply not going to make sense.

When people write of Google AI advantage, they rightly point out the massive data Google has collected over the years. As Google builds its AI models, it has data to test and validate them. IBM got into that trap also and it has to acquire a healthcare company to get data to support Watson Health, its health AI system.

Watson Health today announced plans to acquire Truven Health Analytics, a leading provider of cloud-based healthcare data, analytics and insights for $2.6 billion. Truven will bring more than 8,500 clients, including U.S. federal and state government agencies, employers, health plans, hospitals, clinicians and life sciences companies to the IBM Watson Health portfolio.

IBM knew that Watson Health will not make any significant market-ready progress without the data. So, it bought the data to help train and improve whatever it has been doing with Watson Health.

Two Cases from Fasmicro Group

• When I decided to move into agtech, I knew that AI was going to play a role in whatever we needed to do. So while creating Zenvus, I made sure we also had a roadmap to collect data.That has helped our business because if I had focused on the AI, without the farm data, we would not have made progress. Today, we have the rocket ship and also the fuel. .
• As we also unveil a health AI system for Africa in coming weeks, we have also structured the firm to go beyond just the AI. We will collect data to help us serve our clients and partners. Africa does not have quality health datasets and that means the first thing is to think about the fuel, the data, even before we build the rocket ship (the AI).

All Together

As you move into AI, think about the data. Do not put so much effort in any AI business where you do not have the capacity to get data to test and validate the models. While they have datasets online, most times, the datasets are Western and may not really help you in Africa. For instance, most farm datasets online are from large farms, unlike the type of farms we have in Africa. Those are totally irrelevant to the problems we are solving at Zenvus. Models are great, but it is the data that really makes the business. Data, they say, is as important as electricity in this business, and developing a strategy earlier on for data, for your AI startup, will give you an edge.

Register for Tekedia Mini-MBA (Jun 3 - Sep 2, 2024), and join Prof Ndubuisi Ekekwe and our global faculty; click here.