Reading DeepSeek R1's inner monologue is a jolly good time.

"Wait, maybe...
I'm getting confused...
Wait, I need more context...
But it's unclear...
But wait, no..."

Takes 10 min to replicate with Ollama and the "--verbose" switch.
ollama.com/library/deep...

I wish ChatGPT web would not log me out 1-2 times per day. By contrast, Gemini goes weeks without kicking me out.

Are LLMs capable of scheming or not?

Apollo Research thinks so: "When AI goals are misaligned with humans', weird shit happens"

Rohit Krishnan disagrees "because there's no 'they' there"

In practice, does intent even matter? Or just answer quality?

See their arguments below.

How much AI speedup to expect in software engineering?

Not all companies have the same lived experience so far. See this funny take by Daniel Dines, CEO of UI Path on the 20 VC pod:

Started reading "Think like a CTO" by Alan Williamson out of a sense of duty.

So far, it's comprehensive, actionable, and well-articulated!

One gem: "Office politics can be defined as the absence of data that is replaced by emotion."

Consider it. thinklikeacto.com

Great preso on the state of AI Agents in early 2025 by Prof Graham Neubig at NeurIPS.

I recommend the following personas take a listen

• Designers of agents: how the latest research can get continuously integrated

• Users of AI coding agents: how the OpenHands open-source agent tops SWE Bench

"All modern software was woven in the loom of sorrow"
-Grady Booch

That hits close to home

Prediction Markets make a lot more sense to me after this surprisingly educational @a16z pod (sometimes salesy).

Their accuracy depends on:
- # of participants
- Sum of money wagered
- Diversity of information among participants
- Presence of domain experts
- Speed and accuracy of resolution

This is the digital advertising primer I did not know I needed. For me, it linked:

- Economics of online advertising
- Relative power of ad marketplaces and advertisers
- Changes in tech (e.g. Apple's App Tracking Transparency)
- Strategic gameplay between big tech
- Value of first-party data

1/2

Here's some phenomenal communication advice for techies from Mark Porter, the CTO of dbt Labs.

Check it out. It's great, I promise.

He covers:
- Cascading information
- Intentional escalation
- Starting with the point
- Writing to clarify your thoughts
- Roleplaying difficult convos

1/2

This 7-factor software company due diligence framework feels useful. What do you think?

1- Code Security: Protecting sensitive data from threats.

2- Open Source Licensing: Complying with legal terms of OS components.

3- Code Quality: Assessing the complexity and tech debt.

1/3

"40% of respondents named AI as top contributor to technical debt"

says @Accenture's cloud first senior managing director Andy Tate.

Shots fired?

Do you also think of GenAI as giving starter code that humans have to assess for fitness of purpose and the right standards?

1/2

The one constant in a fast-changing world is human nature.

In the field of investing, Dr. Bryan Taylor and his team went back to the 1600s to understand what affected the value of assets back before we had organized financial markets. Guess what? Same as today: fear and greed.

1/2