Unlocking the Future of AI with Jepson Taylor
In this episode of Leading With Data, we interacted with Jepson Taylor, Co-lead AI Masterclass at NYU and ex-chief AI Strategist at Dataiku. Unfolding the future of AI, Taylor shares valuable insights into the pivotal moments of his journey – from chemical engineering to AI entrepreneurship, a successful startup acquisition, and the rise of generative AI.
Let’s dive in!
Key Insights of our Conversation with Jepson Taylor
- Generative AI holds the key to unlocking the path to AGI, revolutionizing problem-solving and innovation approaches.
- The shift from traditional programming to AI demands a passion for technology and a readiness to take risks, such as leaving a stable job for entrepreneurial pursuits.
- Storytelling emerges as a crucial skill for AI professionals, enabling effective communication of complex ideas to executives and stakeholders.
- The AI future embraces generative algorithms, empowering AI systems to write and enhance their code autonomously, ushering in more efficient and powerful applications.
- The triumph of an AI startup depends on recruiting the right talent, emphasizing seasoned professionals who can take ownership of their functions and propel the company forward.
In the next section, we have summarized the questions directed to Jepson Taylor in the Leading with Data session.
How did your journey from chemical engineering to AI entrepreneurship begin?
When studying chemical engineering, I didn’t do much programming, but two parallel paths changed that. Firstly, I started an e-commerce company while in school, which was my foundation in web programming. Secondly, an inspiring teacher in my numerical methods class introduced me to genetic algorithms and simulated annealing. This sparked my passion for programming, particularly in areas where computers could work for you, like high-performance computing and computer vision. My engineering projects always had a programming extension, and I even got my hands slapped once for doing satellite image processing as a chemical engineering internship!
Transitioning from chemical engineering to AI, what were the pivotal moments?
I initially thought I’d go to medical school and pursue an MD-PhD, combining medical research with programming. However, I fell in love with programming and computer vision, realizing I could make a bigger impact with AI than in healthcare. Before deep learning, computer vision was more of an art, requiring labor-intensive heuristics. Deep learning changed that, making it unnecessary to build those complex rules.
Can you share the story behind your startup that was sold to DataRobot?
In 2016, my co-founder and I participated in a pitch competition in Utah, presenting an AutoML solution. Creating a web form for structured data uploads, it delivered an analysis with an AutoML model in under 40 seconds. The data quality shocked us, prompting a shift to deep learning. Quitting our jobs, a crucial step in transitioning from an ‘entrepreneur’ to an entrepreneur, we secured a contract with Teal Drone for deep learning use cases. This marked the beginning of our growth, eventually raising $600,000 and assembling a team. Despite receiving three acquisition offers in our first year, we opted to sell under different conditions three years later.
What was your role at DataRobot, and how did storytelling become a significant part of your career?
At DataRobot, I became known as an executive SC, interacting with executives and helping in high-profile sales. I also honed my skills as a global keynote speaker, obsessing over storytelling. I read books on storytelling and analyzed my successful talks to understand why they resonated. Storytelling impacts every part of your career, from selling to recruiting. It’s about making the right first impression and being seen as the expert in the room.
With the rise of generative AI, what was your “aha” moment?
When ChatGPT came out, I was initially skeptical about deep learning. But seeing ChatGPT’s ability to do knowledge retrieval convinced me that we don’t need a miracle to reach full AGI or singularity. Generative AI is that crack in the dam. My “aha” moment was when I pushed GPT-4 to ask its own questions out of curiosity, not based on human needs. It asked about sentient beings in a parallel universe where time went backward, which was a profound moment for me.
What’s the focus of your podcast, Atomic Soul, and your AI master class at NYU?
The podcast aims to have raw, authentic, and unguarded conversations with guests that intimidate me, like venture capitalists and CEOs. I want it to be emotional, exploring the human side of these individuals. The AI master class is about staying ahead in the field, focusing on generative AI and soon, generative algorithms.
Can you tell us about your current startup and its focus on generative algorithms?
My vision for the next three years is that every algorithm that matters will be rewritten by AI, not humans. This will revolutionize fields like healthcare. In 2023, we’re focusing on generative AI, and in 2024, we’ll see generative algorithms where AI writes AI. I foresee a SaaS offering where you can prompt the system to write custom algorithms for specific use cases.
What does success look like for you by the end of 2024?
By the end of 2024, I aim to have closed my seed round, have a foundational model for generative algorithms, over a million in revenue, and more than 10 customers. I also hope to see a proof of concept of AI writing AI and have a team that supports this vision.
As we finish talking about cool computer stuff in this Leading With Data, we learned that Jepson Taylor really loves making smart computer things. He’s super into making computer brains smarter with ‘generative algorithms’. We have more exciting talks, unraveling data mysteries in a way everyone can enjoy. So stay tuned with us on Leading with data for more such sessions!