dionlim

[dionlim @dionlim]

@markiewagner @PoeticHQ Congrats to @markiewagner and the Poetic team! Delivering real reliability and real outcomes!

[Markie Wagner @markiewagner]

Introducing @PoeticHQ: a new AI system that executes complex multi-hour tasks with 99%+ accuracy and 10x fewer tokens than agents.

We raised $50M at $500M from Kleiner Perkins, Founders Fund, First Harmonic, and Genius Ventures to build AI that does complex work inside Fortune 500 companies without hallucination.

While code is too brittle, agents are too unpredictable. The work that runs the global economy - anti-money laundering, fraud investigations, underwriting - needs extreme accuracy.

So we built a new kind of software that pairs the flexibility of AI with the predictability of code.

When the world stays the same, Poetic runs fixed code: fast, cheap, identical every time. When the world changes, Poetic uses AI to regenerate its approach and find its way back to the objective.

In one year, we went from zero to an eight-figure run rate as a team of four.

Since then, we’ve scaled the team and executed the highest-stakes processes at AIG, SoFi, and Chime. At SoFi, a large US bank, Poetic reached 99%+ quality on fraud investigations in five weeks.

[dionlim @dionlim]

One thing I love about substack is being able to add in a new thought to an old post. I had a great chat with friend about the UI/UX white space vis-a-vis the labs and added it to my First Mile/Last Mile post. Here is the addition…

A Plethora of Workflow SKUs

Another white space for companies is not only customized workflows, but customized UI/UX interfaces. You can expect the labs to keep their SKUs to a minimum -- e.g., Anthropic’s Claude, Claude Code, Co-Work, Skills, API, etc. These products are accessed through primitives and extensions that enable companies (including the labs themselves) to build their own custom workflows and agents.

While OpenAI and Anthropic move upstream to the application layer and build customized agents, what you are less likely to see is them building customized UI/UX workflows that need to be maintained and supported like an evolving product.

Highly deterministic workflows that are outcome-oriented and labor-intensive but repeatable will be the first fruits of customized agents from the labs. Highly specialized and optimized UI/UX designed to enhance human productivity will be much, much lower on their priority list.

Why? Too many SKUs requiring too much support. Here’s the simple test: if the output can be generated by the agent itself, watch out. But if the tool is custom-designed around making a specific person -- a sales rep, a lawyer, a designer -- ten times more productive, and it needs to evolve with the times, that is a SKU that will be safe from the foundational models.

https://t.co/12YbGvDfM5

[dionlim @dionlim]

The Two Questions Everyone Is Asking

As the labs parade new apps and products weekly, entrepreneurs and investors alike are asking: “Where is the white space vis-à-vis the model companies?” For ordinary people the question is: “What kind of jobs are safe from AI?”

The answer to both those questions is...

[ceodinner @ceodinner]

https://t.co/JMZ4p8yk1v

[ceodinner @ceodinner]

https://t.co/JMZ4p8yk1v

[dionlim @dionlim]

This month's hot topic du jour at our CEO Dinner Insights Report... is software dead? https://t.co/W0jzVemZLO

[dionlim @dionlim]

IQ and EQ Aren't Enough Anymore. The Age of AI Demands AQ.
Intelligence and social savvy are necessary but not sufficient. What separates those who ship from those who stall is having high agency (AQ).
https://t.co/eKGkQAJXPp

[dionlim @dionlim]

If only people had read our November CEO Dinner Insights Report... https://t.co/jvgpsd5tHC

[Peter Girnus 🦅 @gothburz]

https://t.co/2xGooRW1Jv

[dionlim @dionlim]

@dharmesh Love the concept of high CQ. Would love to see you break it down further. E.g., size of context window size x speed/quality of context discernment. "The smartest person in the room is useless if they just walked in." is pure gold.

[dionlim @dionlim]

I've heard for the last 16 years that people would love to be a fly on the wall of our CEO Dinners. With an abundance mindset that is why I started publishing a Chatham House rules compliant CEO Dinner Insights Report.

CEO Dinner Insights: January 2026 https://t.co/f8W5tmGSff

[dionlim @dionlim]

So proud of @itsalfredw and the @ListenLabs team for truly creating a product that is better, faster and cheaper than the time-intensive and labor-intensive process of getting user feedback. A process many companies skip due to the friction. https://t.co/2ABmMGDKOF

[dionlim @dionlim]

CEO Dinner Insights: December 2025
Claude, Gemini and ChatGPT Grade Last Year’s Predictions and Place Odds on 2026 https://t.co/mt2y3MVILD

[dionlim @dionlim]

“These companies have survived for so long on switching costs. You need 50 sales engineers to customize Workday for your organization. AI is going to decimate that.” https://t.co/smB4mvvIgo

[dionlim @dionlim]

“Robin Hood is crypto native and retail friendly. Coinbase is neither.” → Platform positioning and user alignment matter more than first-mover advantages in fintech. https://t.co/l6B1S1it5g

[dionlim @dionlim]

“The amount of money going into depreciating hardware with high margins is the exact same story as Cisco. Somebody will find a way to eat that margin.” https://t.co/smB4mvvIgo

[dionlim @dionlim]

“Neo-labs are all just tweaks on the margin. You need distribution, a product, and a boatload of compute.” → Model architecture innovations alone cannot overcome structural advantages of established players. https://t.co/rcj3mthWT1

[dionlim @dionlim]

“DoorDash faces Zipline doing 3% of their business in Dallas alone with drones. They can deliver a cup of coffee to your doorstep in 15 minutes.” → Labor elimination in delivery creates cost structures traditional players cannot match. #ceodinner https://t.co/rcj3mthWT1

[dionlim @dionlim]

“If you build a great business, no one can hurt you. If you build a bad business, no one can help you.” → External support matters far less than fundamental execution quality. #CEODinner https://t.co/jvgpsd5tHC

[dionlim @dionlim]

“The scarcest resource in Silicon Valley are people (bangers) who can produce new things (miracles) consistently.” → Talent differentiation accelerates as AI handles routine work, making miracle workers 100x more valuable. #CEODinner https://t.co/jvgpsd5tHC

[dionlim @dionlim]

This month’s fun CEO Dinner hosted by @dickc was a throwback format from the old All Things D conference. Name one company you are long and one you are short. And why. Enjoy! https://t.co/HE4rMPmom3