Babak Hodjat is CEO and co-founder of San Francisco-based Sentient Technologies. He was one of the team responsible for developing the core technology that went on to become Apple’s Siri.
To date, Sentient has raised over $170mn in financing and developed the world’s largest distributed AI platform. This platform is being leveraged across several industries including finance where Sentient’s AI autonomously runs its own hedge fund, and ecommerce where AI powers on-the-fly personalization and conversion optimization for major retailers such as Sunglass Hut and Skechers. Sentient has also delved into healthcare, where it has been working on early Sepsis diagnosis, and agriculture, which has seen the company partner with with MIT’s openAg initiative on optimized food growth.
With AI technology increasingly under scrutiny thanks to some high-profile advances in the industry and continued question marks over its deployment, Business Chief sat down with the Sentient CEO to find out more about his plans for the future and the challenges of running a major AI business.
Tell us more about Sentient, your platforms and where you’re at today.
Sentient is an artificial intelligence company and we have built a platform that uses various AI techniques including evolutionary computation, deep learning and rule-based systems. The platform brings those techniques together in a framework that allows modelling, creativity and adaptivity. This is called the LEAF evolutionary AI framework. We’re a product company and currently support two products.
Sentient has clearly evolved rapidly but where do your roots lie and where is the business moving?
Sentient started off building a hedge fund that is AI-powered by LEAF core technology. It launched back in 2016 and established a pretty good life track record – and it’s growing. It’s now open for investments into the fund. For the past two years or so, Sentient has been focused on digital marketing and full-funnel AI enablement. When we think of the digital marketing industry, consumers are coming in through ads and outreached to through a journey typically online or on mobile. That journey is a mix of exploration and exploitation.
These customers want to know what services are available, or they might have in mind a particular product they’re looking for, and then that leads to some form of conversion. For example, buying a product or signing up for a service. Or not. At this point, the business might want to remarket back to them to get them back. That whole journey is orchestrated by the LEAF core platform.
What are the key challenges of running an AI-focused business, specifically regarding expectations versus reality and the hype that tends to surround AI systems and products?
It’s a difficult one, and there are many, but typically it’s around posing a problem and expecting AI to come up with a solution. Within that, there are challenges around actually posing the problem correctly and how to use the technology, which is quite complex in itself. There’s also a challenge regarding how you market and promote yourself: AI has previously over-promised and under-delivered. When you tell someone that your system is AI they expect it to do things that are almost on the verge of mystical and magical!
What you want to do is sell the product on its own merit versus the fact that it is using this advanced technology in the background – that’s one of the biggest issues. For example, in trading you want the track record to speak for itself and in the case of digital marketing it is all about the case study and conversion rate improvement. A positive and negative aspect of AI hype these days is the fact that many businesses are seeking AI solutions and enablement without quite understanding what it actually is. So, AI-related companies are in a position now where they are spending time defining what AI is first and then selling that, which is pretty unique.
Looking ahead, what are your future plans and hopes for the business? More generally, what do you expect to see in terms of AI advancements?
Digital marketing is a large domain and there is coherence in that journey for the customer, from engagement to conversion, and there are many different pieces of software that come into play, all adjacent and all standing to benefit from knowledge discovered in one area. So, even though we started with web optimization and ecommerce and are now working into advertising, there are many other areas that we plan to work on and expand our technology into. The roadmap is there and it affords us the focus and opportunity to present ourselves as we penetrate more into that digital marketing journey. We hope to disrupt the digital marketing world with AI being the centerpiece.
What about your long term hopes for AI deployment in general?
One of my hopes is that we can democratize the use of AI. The current attention on AI is around breakthroughs and the need and availability of data, but there’s one more hurdle to pass before AI can be widely used. That hurdle is to take it out of the hands of the PHDs and specialists and make it easy to deploy in various domains. We believe that everything is an AI problem ultimately. Anywhere you have a decision cycle you can AI enable it. My hope is we can make the creation of AI systems easier.
One of the major discussion points around AI is the question of ethics. Where do you stand on that?
For me, there’s an ethical question around technology generally so I don’t think it’s specifically all about AI. A great example is setting your credit score, which does not have any AI involvement but it’s still something we don’t know or understand. What happens to our private data when we navigate the web? Often AI is not involved in exploiting it, but we don’t understand it so, of course, there are ethics involved there and every day we hear about how they have not been observed.
My general sense is that the ethics of technology are very important and are lagging because technology is advancing at such a rapid pace. It’s our collective responsibility. It’s not just a question for AI scientists and practitioners but humanity as a whole to pay more attention to ethics and the use of technology. By virtue of that, if we solve these problems we won’t have any issue with AI either.
Finally, there have been some very prominent and outspoken critics of AI who fear the potential capability of the technology you’re talking about. There have been accusations that perhaps, given the rapid rate of development within AI in particular, that we’ll find ourselves in a dystopian futuristic scenario that will too easily spiral out of control.
There are many other application areas of AI to be explored. For example, there’s healthcare and agriculture, where we’ve done some interesting research work, and I think those are the areas we should focus on and talk about. Some of these discussions around AI ‘taking over’ and singularity are just noise that is unhelpful because it’s irrelevant. It also tends to come from people who don’t understand the state of the technology and how far away we actually are from some of these science fiction outcomes.