Thread on relevant reflections from Richard Gabriel on AI companies < 1996, with my thoughts:
1/ Today's AI faculty at top schools have 1+ startup. They follow the same patterns: start a co on their open source project, close it, and sell something proprietary & different.
1/ Today's AI faculty at top schools have 1+ startup. They follow the same patterns: start a co on their open source project, close it, and sell something proprietary & different.
2/ Substitute Wall Street for Big Tech here. People evangelize AI use cases & problems at Google / FB / Microsoft. I can't count the number of times I have heard "the future is in training & deploying large models." Yeah well, that future is only for co's with lots of data.
3/ Lack of plans today: container images for ML training & deployment are massive, Co's are building tools using specialized hardware & forcing clients to move to cloud, and all around, ambitious people invent their own abstractions to the same problems, contributing to confusion
4/ The book is "Patterns of Software," which I enjoyed for many reasons as an aspiring writer, history nerd, and design & architecture enthusiast. Richard P. Gabriel (you may know him as a Lisp guy or the "worse is better" guy) wrote it.
Loading suggestions...