iiinova — where your digital self keeps evolving

charles fisher @charleskfisher.bsky.social · 02/04/2025

Muting and blocking. I make a hole in my feed. Refilled with garbage. My haiku for the social algorithm.

charles fisher @charleskfisher.bsky.social · 14/01/2025

Want to learn how to create your own startup? How to be a **founder** rather than a **manager**? Check out my new book "The Relentless Startup" on Amazon!

110

charles fisher @charleskfisher.bsky.social · 30/12/2024

AGI will be underwhelming. An essay and dialogue with ChatGPT: www.charleskfisher.com/blog/agi-wil...

charleskfisher.com

AGI will be underwhelming — Charles K. Fisher

By Charles Fisher and ChatGPT

010

charles fisher @charleskfisher.bsky.social · 27/12/2024

Today is my 39th birthday. During this, my 39th year of life, I lost an uncle, then my grandfather, then my father. Amidst all of that, I stepped down from running the company I started over 7 years ago. In spite of these big changes, I’m the happiest I’ve been in a long time.

240

charles fisher @charleskfisher.bsky.social · 20/12/2024

We have AGI now, but I still need to fill out this paper form and physically mail it back to the bank.

010

charles fisher @charleskfisher.bsky.social · 20/12/2024

I think the hardware lottery is more fundamental than the bitter lesson in machine learning. The hardware lottery causes the bitter lesson. Based on conversations, it seems that not enough people are aware of this paper: arxiv.org/abs/2009.06489

000

charles fisher @charleskfisher.bsky.social · 20/12/2024

Dealing with financial and insurance co's after my dad passed away has shown me, once again, how I live in a tech bubble. Paper mail or phone calls, only. If electronic forms available, they only work on internet explorer or something. Most of America still living in 2008.

010

charles fisher @charleskfisher.bsky.social · 15/12/2024

The Story of the Wise ML Researcher

100

charles fisher @charleskfisher.bsky.social · 06/12/2024

After coding with AI a decent amount in the past few weeks using Cursor, Claude, and now o1, my current assessment is that they are like overconfident undergrad research assistants. They are correct 25% of the time, but think they're correct 100% of the time. But still useful.

000