this week i'm thinking about – #9 – virtual beings

Hey friends,

Happy 2019! I’m writing a weekly blurb about something I learned that’s broadly tech-related.

If you have thoughts, please do reply! I’ll paraphrase the best responses in the next newsletter. My goal is to start conversations with people thinking about similar topics through different lenses. If you know anyone who would be interested in the discussion, please do forward this along or have them subscribe here.


What I learned about: virtual beings 🤖
I saw the Fable Studio demo at Sundance last weekend, which is a Pixar-quality interactive short with a virtual being named Lucy. The experience was captivating, giving me hope that we may be closer than I originally thought to coexistence with virtual beings. I expect we’ve recently crossed the $100M mark of funding into virtual celebrities and avatars between Brud, Fable Studio, AI Foundation, Mica (Magic Leap), Shadows, Artie, and others.

After looking deeper into the industry, I learned about the power of virtual celebrities: specifically in terms of collapsing the agency-celebrity dualism that exists today. There exists a coordination layer between agencies and celebrities, meaning agencies need to enlist new celebrities when new trends or needs emerge. If you’re an influencer, you can’t easily change your persona to adapt to a new trend. But a virtual agency can quickly create new content – or new characters – to fill the gap.

That said, don’t hold your breath for free-roaming interactive virtual friends. Combining 1) NLP-driven understanding, 2) AI-based facial expression, 3) human-quality voice generation, and 4) realistic intelligence will take time to fully realize; we’re currently there on 0 of these 4 key dimensions. The most successful virtual beings of the next decade will likely lead a more confined existence – but entertaining nonetheless.

Top replies from last edition (food delivery, ISAs):
Ayush thinks ISAs are most exciting in the context of third world countries, given their ability to make high quality education more accessible, particularly in a world where remote work is becoming more normalized. He thinks US companies can save 40-60% on talent using a remote work + ISA model. Short-term credit is another key potential use case of ISAs, and could make the loan approval process much more straightforward.

Chris pointed to some players using ISAs in the context of coding schools that have been live for ~4 years. He suggested alternative ISA use cases outside of education such as Uber taking an ISA on future driving income, or truckers splitting their income in a formalized way to pay for ownership of an 18-wheeler.

Matias mentioned a key reason for food delivery market acceleration: food delivery startups successfully transitioned costs from consumers to merchants, which helps ease friction for consumer checkout flows.

Andrew noted the value of food delivery players capturing specific demographic subsegments – e.g. Doordash in dense urban areas like New York, which takes years to get full coverage in.

What I’m reading

  • A 2011 NYT article on "The Facebook Class" at Stanford. Interesting to read in 2019. It makes user acquisition seem trivial: 75 students collectively got 16 million users and $1M in ad revenue in a 10-week period. How often do you see consumer apps getting this type of traction pre-funding in 2019?

  • New dating research from analyzing online dating datasets. Interesting takeaways: 1) people pursue mates that are, on average, 25% more attractive than themselves, and 2) using advanced approach strategies doesn’t meaningfully improve your chances of success.

  • Making 18th-century philosophy cool again: Michael Solana on capitalism.