Aug 06 2019

Neama Dadkhahnikoo

Artificial Intelligence has rapidly accelerated over the last decade. The exponential rise of computational power combined with the wave of big data has yielded some of the most advanced deep learning models to date. We asked you what are your most pressing questions about AI, and we heard you loud and clear. 

Q: Can robots learn from events and develop themselves?

A: That’s an interesting idea because machine learning is actually machines learning by themselves. So, we give them data and they discover patterns and learn behaviors on their own. Additionally, there is another class of AI called reinforcement learning where robots or AI essentially do the same activities over and over and over again and learn from their mistakes. So that’s how researchers teach AI to play chess, to play go, and to navigate mazes. The crazy part is AI gets really good really, really fast. So they learn how to beat humans in chess only after a few hours of training. So yes, robots can learn by themselves.

Q: Where do you see AI in five years?

A: AI has moved so fast in the past five years. The way that its grown the most in my opinion is how ubiquitous it has become in our everyday lives. You don’t know it’s there, but AI is basically in anything you touch that’s electronic, from your phone, to online, to any of the apps you use. They all have AI embedded in them in some shape or form – whether its Gmail learning how you type or it’s a device that can detect objects when you point your camera at it. 

So, now where is AI going in the next five years? Well, some people say it might be self-driving cars. Now that doesn’t mean that all cars will be self-driving, but it might mean that there will be some trucks that are self-driving or there might be cars in specific situations that are self-driving. They can also involve much more advanced technologies in terms of detecting vision, translation, natural language processing, and so on. But the craziest thing is, I think we don’t know where it's going to be in five years, because if you told us five years ago we’d be here now, everyone would say you’re crazy.

Q: What solutions are there for recycling?

A: That’s actually a really great questions because, as part of the AI XPRIZE, we have many teams competing to use AI for Good and to solve the grand challenges facing humanity. And one of those teams is doing “AI for recycling”. What they do is they have built a smart trash can that uses computer vision and looks at the trash you throw away, and also weighs it and other things, and determines whether it is recyclable or not. So, if it’s an apple, it will go to landfill waste but if it’s paper or a cup it will go to the recycling section of the trash can. 

So, computer vision is actually one of the AI technologies you can use for recycling. Another area to think about that does recycling is around blockchain and technologies for where the trash goes. There’s this huge supply chain in terms of recycling, so that the moment you throw trash away there’s this entire industry behind it.  We never think about those behind the scenes parts and there are many areas and opportunities for AI and blockchain technologies to be able to improve those areas as well.

Q: How does one become a cat god?

A: Funny you should ask. I think the way you become a cat god is to spend time with cats, appreciate cats, and really listen to what they need. If they want to be scratched and if they want to be fed, you give them what they want. If you are open and honest and happy with those cats, then maybe you are a god to them.

Q: When will AI be incorporated into the human brain to have things like ready player one?

A: So, first off, Ready Player One was actually VR. The chips weren’t actually in their brain (nerd alert).  Second, the concept of actually having AI inside of your brain comes down to hardware, not software. So we would have to implant electrodes inside of your brain and we do have that technology right now. The problem is that when you put something inside the brain, your body treats it as a foreign object and it immediately builds scar tissue around it and rejects it. So we do have electrodes in brains tech – but it’s just super dangerous and super hard to figure out. Once that technology is solved in some way, I think that AI and technology will be incorporated into our brains very quickly after that.

Q: How does gravity work?

A: Well interesting you ask. I do have a physics minor so I may be able to answer this question. There are four basic forces in physics. We have the strong force, the weak force, electromagnetism, and gravity. Gravity is the only force that works at long distances. Basically when you have two objects, any two objects in the universe, they actually have gravitational forces acting on each other. They pull each other towards one another. For example, when you are on the earth the earth is pulling on you and you are actually pulling the earth up.

Q: So how can AI help solve climate change?

A: Well there are many ways that AI can help with climate change. Anything from better usage of electricity, demand management, and battery storage. So that’s in terms of energy production, energy efficiency, and energy waste. There are also tons of things around biodiversity, deforestation, and conservation that can be done as well. There’s a bunch of interesting things around proactive ways of combating climate change. For example, imagine using AI to manage solar energy in space. To put mirrors into space or making clouds more reflective and using AI to manage it, so less sunlight energy reaches the earth. There are dozens and dozens of different was that AI can actually help with climate change.  The real question is which ones do we want to focus on and which ones do we want to invest in.

Q: What’s the best way to start learning about AI?

A: That’s a good question. AI is such a new field that you don’t have to go to college to study it or to learn about it. There are many courses available online. Whether through things like Corsara or even Stanford University and Berkeley. Many universities open source their courses so you can actually go and take those basic courses about AI online. There are also books that have been published recently from these people like Andrew Ng that actually talk about what AI is. I would also recommend focusing on what you’re trying to do. If you’re trying to code, you might want to learn a language like Python first and then actually start learning about how to code and build those AI models. If you’re someone like a business person or you’re just interested in AI in general, you may want to actually learn more about the core concepts and how AI’s deployed and how AI is used. Because AI is just a tool, so you don’t have to know how to program it to manage It or use it or deploy it inside your technology. Figure out where you want to use those skills and go and read up about those online, in books, and in class as well.


Be sure to follow @XPRIZE on Instagram to take part in our next Q&A. 


Neama Dadkhahnikoo