AI without data would simply not exist.

Garbage in equals garbage out – on the importance of Hi-Fi data

While preparing for HPE Reimagine, I’ve realized that Artificial Intelligence is everywhere. It’s the clothing recommendations when shopping online, it’s smart cars, it’s video games and process automation. It makes life easier for people and businesses alike. And it does that by analyzing data and learning from the data it collects. So, you shouldn’t be surprised when I tell you: the future of technology rests on data.

I could tell you that AI without data would be like Facebook without users but that doesn’t even come close. AI without data would simply not exist. There are three essential parts of AI: the algorithm, computing power and data.

Even if everything else is of the best quality, if you don’t start off with good quality data, you won’t get anywhere. The quality and depth of the data you’ve collected will determine the quality of the AI applications you can build. Even though you may not be starting with AI yet, you should be prepared for a future where your business is only as strong as the data it has collected. Because in a not-so-distant future, what might seem alien to you now, will be embedded in everything we do. And it will all start with the quantity and quality of your data.

You’ve read it well, it’s not just about quantity but about quality. Biased data is one of the biggest challenges when it comes to AI.

It’s quite simple: machines are objective, so if what goes in is biased, what goes out will be too. And even worse, by feeding biased data, we’re developing a vicious circle. A machine doesn’t question the data it receives, it doesn’t realize what bias is because it doesn’t know bias itself. So, it will find correlations that might be wrong and it will keep using those as true. The only way to make sure that what comes out is good, is by feeding it high-fidelity data and we need to start doing that as soon as possible. But how do we achieve that when we’re obviously very biased ourselves? The answer is simple and yet difficult: the time has come to train data analysts not to be biased.

As you can imagine, it requires a lot of energy to fuel the analysis of all this data. The computing power involved in AI cannot be underestimated. And even though that might seem like a big disadvantage for AI, it might not be. Yes, it uses quite a lot of computing power, but it can also reduce the amount of power needed for other processes. Think of

Nest, the home thermostat that adjusts your home temperature once it’s learned your habits. And that’s on a small scale, imagine what it can mean for big companies. Data centers are constantly being cooled, but what if this only happens when it’s actually necessary? The possibilities are endless.

Join us at Reimagine

You can make your data work for you, but you can’t do it thoughtlessly. Let’s discuss how AI can help your business forward at HPE Reimagine on April 24th in The Egg in Brussels. There are only a few seats left. Register today!