For this Deep Tech Dive, I had the pleasure of sitting down with Alishba Imran, Co-Founder of Voltx and Machine Learning (ML) Developer at Hanson Robotics. Voltx is an ML company developing software to accelerate research and development testing of energy storage devices, including batteries and supercapacitors (SC). Hanson Robotics is an AI and robotics company dedicated to creating socially intelligent machines that enrich the quality of our lives. They are the creators of Sophia, the human-like robot. She is also an incoming ML Engineering Intern at Kindred.ai, where she will be developing more intelligent machines using cutting-edge techniques in reinforcement learning (RL) and imitation learning.
- Drug counterfeits are a huge problem in developing countries, and blockchain technology may be a way to fix that.
- One of the greatest barriers to the global adoption of renewable is the insufficient energy storage capacity batteries have. It can take anywhere from 3 months to a year to do production-level testing.
- There is a lot of interest in applying blockchain technology to fractional homeownership and user authentication.
- The current climate agenda will mean that we will focus more on how cryptocurrencies are being mined.
- Going to college is still an attractive option, even if you have what it takes to go straight into the industry. Why? Because of the opportunity to leave your siloed hometown, that you grew up in, and meet people, your age, from another background that think like you.
“To build something meaningful you need to have a long-term vision. We need more people that are working to develop the skills and knowledge that will compound over time so we can solve these major problems.”
This interview was edited and condensed for clarity.
Can you tell me about yourself?
I’m a 17-year-old from Toronto, CA, and throughout high school I self-taught myself ML and blockchain programming. The world fascinates me and how it works. I’m constantly watching documentaries, reading research papers, diving deeper into world issues, and learning about psychology, philosophy, and history. I believe there’s a lot we can learn by seeking perspective to understand and improve ourselves to push humanity forward.
I get excited about the intersection of significant problems with cutting-edge technology because I don’t want to just be a pure software developer or a pure researcher. Currently, I’m working to tackle problems in machine intelligence for bio-inspired healthcare/rehabilitation devices and energy. I want to use my skills and knowledge to come up with novel ideas and execute them. I wanted to apply my skills to solve important world issues.
17 and in AI, blockchain, and robotics! That’s impressive, can you tell me more about your work in blockchain?
Previously, I worked on a platform using blockchain to track 40 to 50% of medication in developing countries that are counterfeit. In these areas, there isn’t a method to understand where these drugs are being produced, how they get manufactured, who’s distributing them. So how do they get to the end consumer? My blockchain product (built on top of Hyperledger) helped to track and authenticate the entire value chain of a drug, from a manufacturer to a consumer. Part of the products got integrated into work that IBM blockchain is doing for supply chain systems globally. I also interned at TD Bank and worked with IBM and SecureKey to develop a decentralized product using Web 3.0 protocols like Solid and Inrupt to automate know-your-customer (KYC) and client authentication. This platform was integrated into Verified.Me and used by the Big Five Banks in Canada.
Okay, so now what about your work in ML and robotics?
Sophia from Hanson Robotics
At Hanson Robotics, I focused on developed Neuro-Symbolic AI approaches for improving the manipulation and precision of tasks done by Sophia the Robot. Also, I worked with a Professor and Master’s student at San Jose State University to develop a new 3D printed material, design, and RL approach that reduces the cost of prosthetics from $10k to $700. Prosthetics today are expensive and can cost up to $100,000 for myoelectric-controlled ones and $10,000 for manual-controlled ones. With RL and generative grasping convolutional neural network (GG-CNN), we are trying to help amputees use these prosthetics more naturally.
Our research can be expanded to neuromuscular rehabilitation for strokes. Stroke is the 2nd leading cause of death, leaving most people impaired for life from a loss of control on their motor functions. This control can be recovered from exoskeleton devices that exist today. However, exoskeleton devices today don’t make it out of the lab and can be costly. I’m working on a project to improve these using soft robotic actuators and RL for control and sensing.
Why did you start Volt X?
I and my co-founder, Shagun Maheshwari, are passionate about climate change issues. One of the greatest barriers to the global adoption of renewable energy solutions is the insufficient energy storage capacity that batteries have when the sun isn’t shining or the wind isn’t blowing. Diving deeper into this problem, we started to realize a common theme; the testing process of batteries before they’re put into production takes a long time. Depending on the manufacturer, it can take anywhere from 3 months to a year to do production-level testing.
If you are constructing a battery for an electric vehicle (EV), you need to know how long it will last and if it will have enough power to withstand harsh conditions. After speaking to manufacturers, we realized that the manual data collection, monitoring, and testing process was a huge bottleneck to more R&D and production of batteries and SCs. There’s a tremendous opportunity here to apply time-series sequence models to help reduce the time of testing and automate redundant manual tasks through data that these companies are already collected from previous tests.
We built the MVP by ourselves while working with a large manufacturer in the SC space. Right now, we are working to improve the product and pilot it with other companies in the space. But this is the start, there’s a lot more to do. Along the way, we realized this problem exists in other electronics, like photovoltaics and LEDs, so we hope to expand our work into those fields one day.
Let’s bring it back to blockchain, I’m curious about where you think this industry is heading this decade?
Within enterprise blockchain, there are many interesting developments in finance, regarding peer-to-peer (P2P) networking and real estate. With P2P, you can direct transactions instead of having to go through a bank. Decentralized finance (DeFi) and open protocols like Uniswap have democratized access to capital markets and will continue to do so in other industries. Digital identity management and Web 3.0 protocols will allow us to create universal wallets and new ways to verify people.
For real estate, fractional home ownership will become a more frequent option. If you think about homeownership today, you only have 2 options to buy or rent a home. But what if you can’t afford to buy and still want to build equity? There are blockchain models that use smart contracts to get you fractional homeownership of a house. These contracts can do title ownership, which can be transferred and regulated easily.
On batteries and EVs, we need more companies that are vertically integrating the entire supply chain process. From mining the materials to manufacturing the battery. There’s no streamlined process for that today. I believe blockchain could help make that happen, by authenticating products and making them easier to track. In the following 10 years, we’re gonna see more EV companies start to accept crypto payments, like Tesla, in their own mode of financing.
As the CTO of a fintech company (Sencha Credit), I can’t ignore DeFi and the transition to web3.0. What do you think about decentralization use cases in the authentication of one’s identity?
Yeah, digital identity is an issue for people who don’t have passports or certifications that can authenticate their identity. You could put all those documents on a blockchain system, like an international wallet. For that, they’ll have to build internal authentication methods that are easily integrable into the frameworks that already exist, which might be a barrier to adoption.
What do you think about mining cryptocurrencies and the effect on the environment?
There are a lot of negative implications for mining, it consumes a lot of energy because of the complex series of algorithms that you need to solve. In fact, it consumes more energy per year than the entire nation of Argentina, which is insane. No government/organization tracks where crypto is being mined and where the electricity comes from. Mining rigs often move from place to place depending on where energy is cheapest, making mining hard to track.
So how we can make mining more sustainable while still maintaining a lot of the underlying principles of blockchain? This will likely mean that we have to change the currency’s algorithm to make its mining more environmentally friendly. Proof-of-stake (PoS) also guarantees that less electricity will be used to mine the currency. Currently, Ethereum’s mining works similarly to Bitcoin’s. Where the most powerful computers can compute a transaction faster so they will have an advantage in getting the most Bitcoin. Ethereum is working on changing the system so miners enter a pool and are randomly selected to complete the transaction and can receive Ether in return. This can help create some change, but overall I think that the environmental impact is likely to grow.
What other industries are you excited about that have the opportunity to be adopted by the public right now?
- Biotech, especially research to accelerate immunotherapies that enhance the human immune system.
- As I said earlier, the R&D process for batteries is prolonged from how long it takes to discover materials. Most materials used today are similar, with about 2 or 3 cathodes and anodes. Computational approaches like AI or quantum computing simulations will be huge with identifying chemistries, and new electrodes or electrolytes for greater energy density.
- Existing AI is still far away from remedying our brain, so AI develops more intelligent robots. There’s research around artificial general intelligence (AGI), better ways to replicate the way the human brain works, from how we process information to making decisions. I’m excited to follow companies like Vicarious that are applying a neuroscience approach using generative probabilistic models. They recently published a paper on how higher-order sequence learning to use structured graphs can support our understanding of how the hippocampus learns cognitive maps.
Most of these things may not be accomplished for 10–15 years, so the mindset of quick returns needs to be refined. To build something meaningful, you need to have a long-term vision. We need more people that are working to develop the skills and knowledge that will compound over time, so we can solve these major problems.
What do you think about the duality of the hype? Obviously, there’s the bad side, the noise that pulls away from the admirable work being done, and there’s also the advantageous side where the hype keeps people interested.
It doesn’t make sense if you are interested in something due to hype. It’s easy to build another social media app, and that’s fine if that’s what you enjoy doing. But we need less of that, and we need more people with long-term thinking working on Deep Tech problems. A long-term mindset helps you focus on what the future will be like and create a transformational change rather than an incremental change. Most are on 2 ends of the spectrum, they are so focused on short-term returns or too focused on long-term thinking where they lack urgency on immediate results. What we need is a bit of both. This is why developing macro and micro-level thinking in everything I do is a key priority for me.
If you look at what’s going on with blockchain right now, including non-fungible tokens (NFTs) and Bitcoin, it’s attractive to the public. To me, this is exciting because it leads to widespread adoption, more people are working on these matters, and more capital is invested. But if you’re working on this for the hype, that’s not dependable. Hype dies out, but the underlying vision/purpose remains. So here, it would be important to believe in the future that DeFi and blockchain can create.
At such a young age, you’ve already accomplished so much, so are you still interested in going to college?
The baseline is to at least experience college, but if I have a great opportunity, I would consider a gap year or drop out. The startup culture promotes founders to drop out, to focus on their company. That’s the culture that we hype up a lot, but going to college can still be a useful experience for 2 reasons:
- Mainly because of the community and people that you meet. Early in your life, you’re very constrained to where you live, the demographic, and the community that you’re around. You’re siloed in your thinking and what is possible. When you go to university, there’s an opportunity to meet people from different backgrounds with unique life experiences. It’s beautiful because you’re meeting people you can learn and grow with.
- The professors and resources can help accelerate growth. The actual curriculum part for undergrad is not as important because a lot of it can be self-taught if you have the discipline to learn and structure your own learning. At the same time, a lot of technical careers require you to have a deep understanding of theory (math, science) — in which these classes can be valuable.