Mark Bennett @mpbennett

Improving decision making, innovation, and productivity through Systems Engineering + RL, Causal, Agentic, Physical Al https://t.co/sQlaWQlkWl talentedapps.wordpress.com Bay Area, CA Joined June 2007

Tweets

3K
Followers

1K
Following

3K
Likes

5K

Aaron Levie @levie

4 weeks ago

Yes, you should still study computer science. The way to get the most out of AI is to have a base understanding of your craft. AI then is a force multiplier in what you can accomplish. This applies to every field whether you're a doctor, lawyer, artist, marketer, or engineer.

a16z @a16z

4 weeks ago

15 35 208 108K 104

Download Video

22 111 451 83K 111

Raul Junco @RaulJuncoV

2 months ago

Code quality didn’t happen by accident. Moments That Defined Code Quality Forever. 1987 – Liskov Substitution Principle Subtypes should work anywhere their parent types are expected. 1988 – Open-Closed Principle Code should be easy to extend without changing existing behavior.…

5 34 222 19K 329

Download Image

Mark Bennett @mpbennett

2 months ago

Before setting policy, consider both Yankelovich’s “McNamara fallacy” and Kerr’s “On the Folly of Rewarding A, While Hoping for B.” Focusing only on metrics while ignoring important qualitative factors results in failure. Misaligned incentives on top of that make it worse.

0 0 0 74 0

Matt Beane @mattbeane

3 months ago

A *great* potential move that could help with skilling/deskilling related to genAI. Eager to test it out.

Olivia Moore @omooretweets

3 months ago

A *great* potential move that could help with skilling/deskilling related to genAI. Eager to test it out.

207 567 10K 830K 3K

Download Image

0 2 7 639 2

Sauers @Sauers_

5 months ago

"When we detect unauthorized crawling, rather than blocking the request, we will link to a series of AI-generated pages that are convincing enough to entice a crawler to traverse them. . . . No real human would go four links deep into a maze of AI-generated nonsense. Any visitor…

59 196 3K 247K 605

Download Image

Mark Bennett @mpbennett

7 months ago

Good demonstration of how to recognize and refute “resulting”

Work Chronicles @_workchronicles

7 months ago

Good demonstration of how to recognize and refute “resulting”

8 318 2K 70K 274

Download Image

0 0 0 107 0

Andrej Karpathy @karpathy

8 months ago

"Move 37" is the word-of-day - it's when an AI, trained via the trial-and-error process of reinforcement learning, discovers actions that are new, surprising, and secretly brilliant even to expert humans. It is a magical, just slightly unnerving, emergent phenomenon only…

443 1K 10K 991K 4K

Andrej Karpathy @karpathy

8 months ago

For friends of open source: imo the highest leverage thing you can do is help construct a high diversity of RL environments that help elicit LLM cognitive strategies. To build a gym of sorts. This is a highly parallelizable task, which favors a large community of collaborators.

322 845 9K 1.2M 3K

Adam Grant @AdamMGrant

8 months ago

Hiding knowledge from others doesn’t just undermine them. It limits your thinking too. People who withhold information end up having less creative ideas. Their peers don't trust them enough to share insights with them. A key to learning is being generous in teaching.

41 542 2K 124K 518

Download Image

Matt Beane @mattbeane

9 months ago

Amazing to watch professionals "discover" my findings as they deal with LLMs. Steve's last post caught fire, centered on my "novice optional" finding from 2015. Now he asks who *is* building new skill? Answer, roughly, is my "shadow learners" from 2018. sourcegraph.com/blog/the-death…

1 3 18 1K 11

Henrik Karlsson @phokarlsson

2 years ago

Graph of decision quality among professional Go players. A sudden increase after AlphaGo. It is not only because they are learning from the AI. Players are suddenly inventing new moves at a faster rate too!

54 485 3K 766K 843

Download Image

Levi @levikul09

2 years ago

Have you ever seen data that contradicts your expectations? It could be due to Simpson's Paradox. Trends in different groups can reverse when combined. Let me explain this phenomenon. 🔽…

12 175 977 257K 600

Download Image

Cameron R. Wolfe, Ph.D. @cwolferesearch

2 years ago

Given the current foundation model paradigm, I wonder if building/training models will become antiquated. Will future data scientists understand the details of optimization, architectures, etc.? ML may slowly be abstracted in favor of simpler (language model-based) solutions...…

19 86 630 205K 512

Download Image

Mike Lazarus @mikelazarus87

2 years ago

🧵1/13: One week ago, I reached out to 14 ai experts and asked them ONE question: 👉"What Was Your "Aha" Moment That Made Your Realize The Potential Of Ai?" 🔥To my surprise, 11 of them answered, so I compiled all of their answers here in this thread (hint: #9 left everyone🤯)