Why AI is an opportunity not a threat, for businesses
Jose Berengueres, Associate Professor at UAE University, answers the question that is currently on everyone’s mind…
After 20 years in the doldrums, the art of making machines think for us is experiencing a renaissance. Two decades of drops in the cost of the computing power of Nvidia cards have enabled machines to finally be better than humans at some specific tasks such as classifying cats on YouTube, recommending Netflix movies and programming cars to drive. However, it does not stop there. In five years, robots will better than us at things from diagnosing cancer to babysitting kids. And our laptops will have IQ equivalent to 25 points. So naturally, things will change a lot. Some companies are already taking steps to prepare for this up and coming revolution. One example is Uber. A company currently valued at 70bn and has negative cash flow. Three years ago, Uber woke up to this AI epiphany and sent someone to the robotics department at Carnegie Melon University (one of the best in the world in robotics). Three months later, Uber had recruited the whole department to work in their newly formed self-driving division. Of course, not everyone at CMU was happy about that because the department became deserted and that created some PR issues. Uber later made a $100m donation to CMU. Guilt apart, Uber knows that their current business model is not sustainable in the end. Uber has the best transport app available today, but the problem is that users don’t mind switching to a rival app (such as Lyft). In other words, the Uber app is not as sticky as the Facebook or Instagram app. As sad as it sounds, the only way forward to keep its customers in the future is to be the cost leader and human drivers are in the way of that vision. And that is why they are betting the whole company strategy on developing self-driving tech. So far, compared to Tesla it is not going well, but the jury is still out. This, and hundreds of other examples (See Google vs. Levandosky case and others…) have exposed the hunger for AI talent. This is good for us as it brings funding, industry collaboration and attracts talent to the field. To give a sense of how competitive the field has become just note that the top four AI academics in the world: Geoff Hinton, Andrew Ng, Yan LeCun and FeiFei Li all have been recruited by Fortune 500 companies.
This year, the former Stanford professor Andrew Ng just tweeted “the opportunities in #AI field are larger than the talent available”. And it is true. Colleagues are leaving academia and joining transnational companies thanks to sweet salaries, exciting challenges and excellent computing infrastructure that universities usually cannot afford. For example, Google Deep Mind spends the equivalent of AED 300 to 1,000 million on computing credits per year. However, there are downsides too. Bright students are being snatched up too. One of the students I mentor scored a world record at Atari Pac-Man game challenge hosted in August by www.open.ai, a website that is the equivalent of the Olympics of AI. Four days later, he had been hired by {sourced.tech}, an AI start-up based in Madrid.
However, although a lot is being written about AI, I think there is still a lack of understanding among SME business owners as to how they can fully benefit from this revolution. The number one advice is that you don’t need to be Amazon or Uber to take advantage of the coming AI wave. As Michio Kaku said, AI will be available on tap. Just like electricity is today. The second advice is to realise that data is the new oil. For example, Facebook makes US$ 10 or revenue per user per year. All of it is thanks to the AI that makes Facebook one of the most targeted and effective ads systems in the World. Facebook is already an AI-driven company, but it would be nothing without the data. Same goes for Airbnb, booking.com, Tesla and Google. So, if you have data and you are not squeezing every single penny out of it, you are missing out revenue stream and you will die because some competitor will do it. The last lesson can be learned from the revolution that started with AOL in 1996 and ended with the dot-com bust of 1999. Some businesses saw the internet coming and embraced it, others ignored and now they are zombies. Here’s a token that will sound familiar: “Technology by itself is not the real disruptor. Being non-customer centric is the biggest threat to any business.”
Having said that, the refusal to embrace the use of AI today is like refusing PC in the office in 1999. It will hurt you at the end. Here at UAEU, we partner with forward-looking companies and start-ups that get this. Together we analyse data and develop predictive models, what is also known as predictive analytics. It is a win-win. We don’t have data but know how to make it useful. Companies have data but the knowledge to make it useful in some cases is so new that is not found inside them yet. Working with real data allows us to advance the field and publish papers. Companies get a state of the art AI that keeps them ahead of the competition.
In the case of an impresario that wants to learn more about how to use AI here in UAE, where can she go? There are a few venues to keep abreast the latest developments in the field. One of the venues I use to recruit talent is the Dubai Data Science meetup and the PyData meetups worldwide. Both cutting-edge and hands on.
