Ishan Pandey: Hi Kenso, welcome to our series “Behind the Startup.” Please tell us about yourself and the story behind Morphware?
Kenso Trabing: I moved to India in 2017 to open up a data science boot camp on behalf of a FinTech-focused coding boot camp in midtown Manhattan. Part of my job was teaching the first few cohorts, but another was helping to set up the business, with the incredible team we had there, and part of that entailed securing the computational resources that students needed.
I was surprised to learn how expensive it was to rent a GPU, from an infrastructure-as-a-service provider, but the experience was also a little less than ideal because the internet would go out a few times a day and we couldn’t connect to the centralized service.
I learned quite a bit about money in India, too. It was the spring after the government canceled all the large notes and cash was extremely hard to come by. I couldn’t even open a bank account and wires were practically impossible, for some reason, as well. The most surprising thing to me, though, was that one of my brightest students— who had a graduate degree from the London School of Economics— held a remarkably large amount of ether.
I ended up teaching Solidity, the smart contract programming language, in a series of guest lectures at Yale, when I returned from India, but the trouble I had using an infrastructure-as-a-service provider to train machine learning models, while I was in India, never left me. Morphware is an original solution to this problem: it’s a decentralized approach to deep learning.
Ishan Pandey: Please tell us a little bit about Morphware? Further, what are some major issues surrounding machine learning?
Kenso Trabing: The main problem that Morphware is addressing is: the computational complexity required to train a machine learning algorithm doubles approximately every 3.4 months. Moore’s Law, which some argue has been tapering off, posits that the number of transistors in a processor doubles every 18 months. That’s a huge gap, and it’s widening.
The way our program is currently written, worker-nodes bid on the right to run machine learning workloads that data scientists post to the network; in a Vickrey auction process. The second iteration of the protocol will leverage event streams so that multiple machines can work together with one another on a single workload.
Ishan Pandey: DeFi is now widely acknowledged as a significant shift within the financial market ecosystem. Do you think DeFi should be regulated?
Kenso Trabing: I think that DeFi has a lot of room to disintermediate TradFi processes and pass the efficiencies onto consumers in the form of better rates and faster transaction speeds, but there are also a lot of unscrupulous people out there; so I can understand the point of view behind each side of the argument.
Ishan Pandey: Web3 is an attempt to integrate blockchain, smart contracts, and decentralized applications, although its exact utility is difficult to determine. Please elaborate a little on your thoughts about Web 3.0?
Kenso Trabing: I’m a true believer. I honestly believe Web 3 will revolutionize the way people interact with one another, across time and space.
Ishan Pandey: Can you elaborate a little bit on the Solana blockchain ecosystem and its ‘proof-of-stake’ mechanism?
Kenso Trabing: I think the most interesting thing about Solana is the
RAM requirements for the minersIshan Pandey: NFTs are becoming increasingly popular especially with YouTube now hinting at plausible NFT features. What are your views on NFTs and its ever-rising popularity and its future?
Kenso Trabing: I think it’s great that NFTs are being explored as a medium to circumvent the classical power structures in the contemporary art world, but I think that it’ll be really cool for NFTs to take the place of IRL stock certificates and property deeds. I’ve heard of some cool projects working on stuff related to this, but I can’t remember what they are.
Ishan Pandey: What are your views on GameFi and Play-2-Earn’s future in the digital economy?
Kenso Trabing: A guy on our team, Matt Topping, actually pointed something really interesting out to me about GameFi. I couldn’t believe the level of ingenuity involved here, but apparently, some games are tightly coupled abstractions of DeFi processes; like yield farming. I’m not the best person to ask about it, but that’s something I’m really bullish on: making DeFi more approachable by gamifying the experience.
I think Douglas Hofstadter said that analogy is the core of cognition, and some of the topics in DeFi are fearsomely esoteric; so I’m very bullish on that sub-set of the space.
Ishan Pandey: What does the roadmap ahead look like for the crypto ecosystem as a whole?
Kenso Trabing: The roadmap ahead for us is two-fold. The current state of the system is MVP-like, so we’re working on making that more production-oriented. The later stage of this will be the design and development of a mobile application, so people will be able to earn crypto on their cell phones by running a validator node on the network. The next generation of the system, as I mentioned earlier, will revolve around upgrading the protocol so that multiple machines can work with one another on a single workload.
The roadmap for the crypto ecosystem as a whole seems to revolve around scaling, right now, but there’s obviously a lot going on outside of that, as well. I’m really excited about what’s to come.
Disclaimer: The purpose of this article is to remove informational asymmetry existing today in our digital markets by performing due diligence, asking the right questions, and equipping readers with better opinions to make informed decisions.