Keep pulling the thread on Mark Pincus.
Mark Pincus believes that while the consumer technology sector is currently considered un-investable, the opportunity to create new major consumer services is greater than ever due to AI and agents.
Garry Tan predicts that an always-on, conversational consumer AI assistant will become a company valued at $10 billion or $100 billion.
Mark Pincus argues that there is currently no proven distribution path for consumer technology companies, which makes building in the sector very difficult.
Garry Tan cites a recent statistic indicating that 90% of enterprises that have invested in AI have not yet received any benefit.
Garry Tan states that Peter Steinberger is reportedly spending $1 million to $1.1 million per month on AI tokens.
Garry Tan advocates for a new software development paradigm where developers write Markdown to instruct LLMs to generate code, rather than writing code that calls LLMs.
Garry Tan predicts that the cost of using frontier AI models will decrease by orders of magnitude within two years, from around $1 million per year to as low as $1,000 per year.
Garry Tan predicts that the main consumer revolution in AI will occur around 2029, once the cost of using powerful models decreases by approximately three orders of magnitude.
Garry Tan cites Suhail Doshi and Jeff Dean in predicting a 10,000x increase in AI inference efficiency is likely.
Garry Tan predicts that the next generation of major consumer social companies, comparable to Meta and Snap, will emerge within the next couple of years due to AI.
Mark Pincus states his early social network, Tribe, failed because it did not correctly implement a component of user trust.
Early investors in Facebook, including Mark Pincus and Peter Thiel, sold their stock early because they did not anticipate how large the company would become.