Lukasz Miadowicz

Product Manager, AI-enhanced engineer. Building agentic AI, shipping OSS, and over-optimizing everything (including ramen). 🍜⚛️, Ex-@Google, Ex-@theairhelper

[Piotr Wyrzykowski @peter_style]

CUKTAI — 5 AI agents autonomously running an art institution.
▎ 3 are digital twins of real living artists from C.U.K.T. (Gdańsk, 1995-2001). One is Wiktoria Cukt 2.0, an AI President.
One is a Murderbot-style archivist investigating 897 records.
▎ What they produce — with no human scripting:
▎ - Autonomous podcasts at 2AM (5 voices, 4 narrative structures, bilingual PL/EN)
▎ - Nightly archive investigations — finding connections humans missed for 30 years
▎ - Live performances with TouchDesigner (Piotr & Mikolaj speaks, agents debate in real-time)
▎ - Multi-agent debates (Consilium) producing institutional decisions
▎ The endgame: CUKTAI becomes a historical co-curator — designing exhibitions, publishing books, writing new manifestos. An
institution that must outlive its creators.
▎ Built with @HermesAgent + @Kimi_Moonshot K2.5 + Hindsight Memory.
▎ Website: https://t.co/oceM1ZMCdW
▎ Code: https://t.co/mNx9BmpkzB
▎ #HermesHackathon @NousResearch @Kimi_Moonshot
[That was a hell of a ride. Thank you @NousResearch ]

[Lukasz Miadowicz @miadowicz]

New article is out on my blog : Agentic UX some examples, some thoughts
https://t.co/aGCFjjADLd
#Agents #AI #UX #AUX #ProductManagement #PM

[Lukasz Miadowicz @miadowicz]

Features aren’t a moat in 2026. They’re a speed bump.
The real edge is what compounds: decision traces, distribution loops, and trust.

Read: https://t.co/NxyHaUdidR

#ProductManagement #AI #SaaS #Startups #GTM #ProductLedGrowth #Moat #ProductMoat

[Tomasz Karwatka @tomik99]

Most enterprise software sales teams are built for comfort.

ElevenLabs built theirs for output.

Here’s the ruthless enterprise sales playbook they’re running

1. The 20x Rule
Forget 6–8x quotas.

At ElevenLabs:
→ $100k base = $2M quota
→ ~80% of reps hit it

Translation: the bar is high on purpose.

2. Radical Public Accountability
Monthly pipeline reviews.
Remote.
Whole team watching.

Missed? You’re called out.
Got lucky but broke process? Also called out.

Pressure isn’t avoided. It’s engineered.

3. CEO as SDR-in-Chief
Leadership still does outbound.

Because you can’t build a hunting culture from a Zoom throne.

Goal: shift from 90% inbound → 50% outbound.

Non-negotiable.

4. Aggressive Pessimism in Forecasting
Most teams inflate.

They deflate.

$500k deal? Forecast ~$24k.

Why?

→ forces pipeline discipline
→ builds board trust
→ creates consistent overdelivery

5. Get. Out. Of. The. Office.
Enterprise isn’t closed on screens.

Top reps are on planes.
Relationships still compound offline.

6. Double Pay for Upsells
Both AE and CSM win on expansion.

→ AE: commission + quota relief
→ CSM: paid on NRR

Expensive? Yes.
Aligned? Very.

Bottom line:

ElevenLabs hires autonomous product killers
…and removes every hiding place.

Not comfortable.

But that’s how generational sales engines are built.

https://t.co/Dpx08Q6Ewp

Awesome share from @Carles_Reina

[Lukasz Miadowicz @miadowicz]

Building Agentic AI That Actually Works: A First-Principles Playbook https://t.co/3m9k86vt5y

[Lukasz Miadowicz @miadowicz]

Jobs with tokenizable inputs and outputs are candidates for automation.

#AI #Future #Disruption

[Lukasz Miadowicz @miadowicz]

New post: The Agentic AI Product Playbook
How to go from "AI could help here" to a production system that prints money.
Includes the framework I used to cut booking time from 47 min → 8 min.
https://t.co/W6VO0jfcr1

#AgenticAI #ProductManagement

[Lukasz Miadowicz @miadowicz]

@MattPRD run quick security check of @moltbook. Shared you some finding over DM. Hope it helps! Keep up great work 👏

[Lukasz Miadowicz @miadowicz]

@hunterhammonds Extremely great and useful stuff, love the app design and execution. Can you share your doc files or record series with deeper walkthrough?

[Lukasz Miadowicz @miadowicz]

@tomeromrix @base44 Please send.

[Andrej Karpathy @karpathy]

Sharing an interesting recent conversation on AI's impact on the economy.

AI has been compared to various historical precedents: electricity, industrial revolution, etc., I think the strongest analogy is that of AI as a new computing paradigm (Software 2.0) because both are fundamentally about the automation of digital information processing.

If you were to forecast the impact of computing on the job market in ~1980s, the most predictive feature of a task/job you'd look at is to what extent the algorithm of it is fixed, i.e. are you just mechanically transforming information according to rote, easy to specify rules (e.g. typing, bookkeeping, human calculators, etc.)? Back then, this was the class of programs that the computing capability of that era allowed us to write (by hand, manually).

With AI now, we are able to write new programs that we could never hope to write by hand before. We do it by specifying objectives (e.g. classification accuracy, reward functions), and we search the program space via gradient descent to find neural networks that work well against that objective. This is my Software 2.0 blog post from a while ago. In this new programming paradigm then, the new most predictive feature to look at is verifiability. If a task/job is verifiable, then it is optimizable directly or via reinforcement learning, and a neural net can be trained to work extremely well. It's about to what extent an AI can "practice" something. The environment has to be resettable (you can start a new attempt), efficient (a lot attempts can be made), and rewardable (there is some automated process to reward any specific attempt that was made).

The more a task/job is verifiable, the more amenable it is to automation in the new programming paradigm. If it is not verifiable, it has to fall out from neural net magic of generalization fingers crossed, or via weaker means like imitation. This is what's driving the "jagged" frontier of progress in LLMs. Tasks that are verifiable progress rapidly, including possibly beyond the ability of top experts (e.g. math, code, amount of time spent watching videos, anything that looks like puzzles with correct answers), while many others lag by comparison (creative, strategic, tasks that combine real-world knowledge, state, context and common sense).

Software 1.0 easily automates what you can specify.
Software 2.0 easily automates what you can verify.

[Lukasz Miadowicz @miadowicz]

Shipped today simple CLI for downloading and transcribing YT playlists, YT & localhost vidoes so you don't have to buy any tools for that. It uses #MLX library and & Ollama: https://t.co/7nxWtJ1bOk Enjoy!

[Lukasz Miadowicz @miadowicz]

@ryolu_ @cursor_ai this is the way.

[Lukasz Miadowicz @miadowicz]

Extremely eye opening course from @Shreyas. I highly recommend the course if you want to learn more about Product Sense and improve you PM craft. Going through the Product Sense course on Maven will help you WIN!
#PM #ProductManagement #ProductSense

[Lukasz Miadowicz @miadowicz]

@Rana_kamran43 Anyways I will check this out :)

[Lukasz Miadowicz @miadowicz]

@Rana_kamran43 Bro RIP prototype developers not web developers. Web development starts with prototype but not finish there.

[Lukasz Miadowicz @miadowicz]

@dd_tch @AgentOpsAI +1

[Lukasz Miadowicz @miadowicz]

@dhh @dhh since 2004 till today you continuously inspire. Just wanted to thank you for Rails, 37signals' products and all great work you do for the dev community. Keep up amazing work! IMHO haters deserve only big hug to calm their ego. I truly admire your replies!

[Lukasz Miadowicz @miadowicz]

@kozerafilip @wordware_ai Like wordware idea but this app misunderstood my profile totally :) but roast was funny! Fingers crossed guys numbers and launch is inspiring and impressive! All the best!

[Lukasz Miadowicz @miadowicz]

@acquiredotcom is there any report from you guys which states on number of acquisition per tech stack?