The "T" shape model for Generative AI
What kind of skill sets do individuals, teams, and organizations need in the era of AI?
Programming Note 1: SFTW will be off next week due to the US Independence Day holiday. I will be back in your Inbox on July 9, 2025.
Programming Note 2: It has been a few months since the launch of the SFTW Plus (paid tier). I would love to get some feedback from you about SFTW, irrespective of whether you are a free or paid subscriber. Would you be willing to spend 20-25 minutes on a video call to answer a few questions? I promise to tell you some dad jokes!
The “T” shaped model for GenAI
When I worked at Alphabet X, the moonshot factory, the company looked for people with “T” shaped skills.
In the context of Alphabet X, the "T-shaped talent" refers to individuals who possess a deep expertise in a specific area (represented by the vertical stroke of the “T”) combined with a broad range of knowledge and skills (represented by the horizontal stroke of the “T”) that allow them to collaborate effectively across different disciplines and backgrounds (the horizontal stroke).
Given many projects would die inside the company, it was important that the people hired to work there had the intellectual and emotional capability to shift to a different problem, if necessary.
“Silos are good”
“Silos” is a heavily used trope in all industries.
Information silos.
Teams don’t share information. The left hand does not know what the right hand is doing. It results in sub-optimal results.
Or information is stored such that it is difficult to access.
Professor Hollis Robbins turned the silo trope on its head and wrote another (!!) fantastic essay. She used the example of grain silos (with some fascinating historical background), to illustrate how going deep in a particular area is very beneficial for academic institutions.