Disclaimer: The views and opinions expressed in this blog are entirely my own and do not necessarily reflect the views of my current or any previous employer. This blog may also contain links to other websites or resources. I am not responsible for the content on those external sites or any changes that may occur after the publication of my posts.
End Disclaimer
For obstinacy standing alone is the weakest of all things in one whose mind is not possessed by wisdom.
― Aeschylus, Prometheus Bound
AIML
Will A.I. Be a Bust? A Wall Street Skeptic Rings the Alarm.- GenAI can be transformative and a huge bubble.
“building too much of what the world doesn’t need ‘typically ends badly.’”
Sam Altman: The Intelligence Age-
“In three words: deep learning worked.
In 15 words: deep learning worked, got predictably better with scale, and we dedicated increasing resources to it.”
Cloudflare’s new marketplace will let websites charge AI bots for scraping-
“If you don’t compensate creators one way or another, then they stop creating, and that’s the bit which has to get solved”
Closed-source vs. open-weight models: The Labonne Graphic:
Two kinds of LLM responses: Informational vs Instructional-
“there are two kinds of distinct responses people get from LLMs: informational and instructional.”
“this is more pronounced in the domain-specific RAG systems like that of enterprises where people may forget SOPs for different tasks. People ask about how to get bigger inbox quota, HR policies, how to apply for vacations etc. Things often boil down to a “how-to” question since RAG skips the “where can I find…” step by doing that for you.”
Exponential growth brews 1 million AI models on Hugging Face-
"Contrary to the '1 model to rule them all' fallacy," he wrote, "smaller specialized customized optimized models for your use-case, your domain, your language, your hardware and generally your constraints are better. As a matter of fact, something that few people realize is that there are almost as many models on Hugging Face that are private only to one organization—for companies to build AI privately, specifically for their use-cases."
Painting
William Logsdail (British, 1859–1944), St. Paul's and Ludgate Hill, c.1887
Cyber
Markets
Founder mode vs Bureaucrat mode- Read this
“If winners hire winners, and losers hire losers, what do bureaucrats hire? More bureaucrats of course.”
OpenAI closes in on largest VC round of all time- of. all. time.
“WHAT has the stock market returned over the long run? About 8-10% per year.
But HOW — HOW has the stock market returned 8-10% per year? — is a whole another question.
The stock market is NOT a rising tide lifts all ships story. It is a haystack with unknown, but required, needles story. So buy the damn haystack.”
Real GDP Per Capita Growth in the G7:
Why millions of Americans give up control of their thermostats-
“That’s why programs like this one are called ‘virtual power plants.’ “
It's hard to write code for computers, but it's even harder to write code for humans-
“A succinct list:
Developers getting to success faster are happy developers. They will like your tool.
Developers banging their heads against errors are sad developers. They will blame your tool.”
Misc
Do Animals Know That They Will Die?- “An existential mystery”
Octopuses and fish caught on camera hunting as a team- Pixar- hear me out…
“The other fish provide several options, and then the octopus decides which one to take. There’s this element of shared leadership.”
Why the U.S. Can’t Build Icebreaking Ships- The US has 2 icebreaking ships
“The U.S. has interests in regions all around the globe. Perhaps none are more remote than the polar region: areas near or in the Arctic and Antarctic Circles.”
Introducing granular MTA bus speed data- cool dataset, link is at bottom of page on the NYS Open Data Portal
Podcasts for a Road Trip
Chameleon: Hollywood Con Queen
Books
My Favorite Business Mischief Books
4 books I’ve read and 5 books I want to read