Apollo performed extensive testing and looked at the chain of thought of models like o1. Sure looks like scheming:
www.apolloresearch.ai/research/sch...

But Rohit provides compelling arguments that it's not due to the lack of an "entity":
www.strangeloopcanon.com/p/no-llms-ar...

Strange times

Daniel - "Look, with all respect to Salesforce, our technology is much harder to build. What we are building spans many different technologies. We will get some productivity improvement but I don't think it'll be gigantic."

youtu.be/1WUuppfscXw?...

Kinda mirrors our speedup experience so far, tbh

Host - "Benioff said that software engineer cohorts have been made so much more efficient, they're now using mostly AI tools. How are you seeing efficiency within development teams change internally at UI Path?"

• Builders of websites and APIs: how structured input/output makes life easier or LLMs and agents

References

Preso: www.youtube.com/watch?v=B6PK...

OpenHands: github.com/All-Hands-AI...

SWE Bench www.swebench.com

Got some unique insights about future events? Great place to put your money where your mouth is.

A fun informative listen:
a16z.simplecast.com/episodes/pre...

Plus this banger: "targeted advertising is a public good"

Bravo and thank you to @ericseufert.bsky.social for dropping this knowledge on the Making Markets pod hosted by Eric Golden

open.spotify.com/episode/7HNI...

2/2

You'll also find surprising bits about rent, school fees, parenting and the "FOS" letter.

Hear @marklovestech.bsky.social on the Modern CTO podcast
www.youtube.com/watch?v=-rdq...

2/2

I can't speak to their product, but I appreciated the overview by Matt Van Itallie, CEO of SEMA.

Hear him on @PEFunCast:
www.youtube.com/watch?v=dRUc...

3/3

4- Team Retention: Maintaining a stable and knowledgeable dev team.

5- AI Composition: Evaluating the implications of GenAI in the code.

6- Maintainability: Modifying and updating the codebase.

7- Cybersecurity: Assessing the overall security posture of the software and its environment

2/3

On the IT Visionnaries podcast, he emphasizes seven foundational capabilities for AI readiness:

-infrastructure
-security
-control plane
-data
-AI backbone
-applications
-digital platform

Do you find that a useful framework?

www.youtube.com/watch?v=tSHC...

2/2

#GenAI

If you are around, let's catch up.

Specifically, four factors impact market returns: war, inflation, socialism, and autarchy.

The more things change, the more they stay the same.
Hear more insights from Dr Taylor, Chief Economist for Global Financial Data, on the Rational Reminder podcast:
www.youtube.com/watch?v=_FiC...

2/2