Zoltan Varju
@zoltanv.bsky.social
Data scientist exploring intersections of philosophy, science, and AI. Writing about where technology meets human understanding and statistical thinking. Based in Hungary. Blog: crowintelligence.org
how ancient philosophy created modern mathematics—árpád szabó's revolutionary discovery about the origins of rigorous thinking
mindtheminds.substack.com/p/how-did-ma...
mindtheminds.substack.com/p/how-did-ma...
How Did Mathematics Become a Science?
Árpád Szabó's Revolutionary Discovery
mindtheminds.substack.com
November 13, 2025 at 12:34 PM
how ancient philosophy created modern mathematics—árpád szabó's revolutionary discovery about the origins of rigorous thinking
mindtheminds.substack.com/p/how-did-ma...
mindtheminds.substack.com/p/how-did-ma...
The Turing story you know is incomplete. Polish mathematicians broke Enigma first (1932-39). French intelligence brokered the handoff. Y Service intercepted messages. Turing mechanized it at scale.
New essay on genius + collaboration:
mindtheminds.substack.com/p/beyond-ble...
New essay on genius + collaboration:
mindtheminds.substack.com/p/beyond-ble...
Beyond Bletchley
The Real Story of Alan Turing and the Team That Changed Computing
mindtheminds.substack.com
November 4, 2025 at 1:40 PM
The Turing story you know is incomplete. Polish mathematicians broke Enigma first (1932-39). French intelligence brokered the handoff. Y Service intercepted messages. Turing mechanized it at scale.
New essay on genius + collaboration:
mindtheminds.substack.com/p/beyond-ble...
New essay on genius + collaboration:
mindtheminds.substack.com/p/beyond-ble...
New post: Managing Python Environments with pyenv and uv 🔧
The "works on my machine" problem isn't about your code - it's about environment management.
Part 1 of my series on transitioning from beginner to advanced data science practitioner.
crowintelligence.org/2025/10/09/m...
#DataScience #Python
The "works on my machine" problem isn't about your code - it's about environment management.
Part 1 of my series on transitioning from beginner to advanced data science practitioner.
crowintelligence.org/2025/10/09/m...
#DataScience #Python
Managing Python Environments: pyenv and uv Tutorial (Data Science Engineering Gap Part 1) - Crow Intelligence
You've written Python code that works perfectly on your laptop, but when you share it with a colleague or return to it months later, nothing runs. This comprehensive tutorial teaches you how to manage...
crowintelligence.org
October 9, 2025 at 2:43 PM
New post: Managing Python Environments with pyenv and uv 🔧
The "works on my machine" problem isn't about your code - it's about environment management.
Part 1 of my series on transitioning from beginner to advanced data science practitioner.
crowintelligence.org/2025/10/09/m...
#DataScience #Python
The "works on my machine" problem isn't about your code - it's about environment management.
Part 1 of my series on transitioning from beginner to advanced data science practitioner.
crowintelligence.org/2025/10/09/m...
#DataScience #Python
Yale: 33 months since ChatGPT, no measurable job disruption yet.
Occupational mix changing only slightly faster than with computers/internet—and started before ChatGPT.
Transformative tech takes decades, not months. Exactly the data we need.
budgetlab.yale.edu/research/eva...
Occupational mix changing only slightly faster than with computers/internet—and started before ChatGPT.
Transformative tech takes decades, not months. Exactly the data we need.
budgetlab.yale.edu/research/eva...
Evaluating the Impact of AI on the Labor Market: Current State of Affairs
budgetlab.yale.edu
October 7, 2025 at 9:59 AM
Yale: 33 months since ChatGPT, no measurable job disruption yet.
Occupational mix changing only slightly faster than with computers/internet—and started before ChatGPT.
Transformative tech takes decades, not months. Exactly the data we need.
budgetlab.yale.edu/research/eva...
Occupational mix changing only slightly faster than with computers/internet—and started before ChatGPT.
Transformative tech takes decades, not months. Exactly the data we need.
budgetlab.yale.edu/research/eva...
Philosopher John Searle died last month at 93. His Chinese Room Argument—can computers ever truly understand, or just manipulate symbols?—defined the AI consciousness debate.
With ChatGPT & Claude everywhere, the question feels urgent again.
mindtheminds.substack.com/p/understand...
With ChatGPT & Claude everywhere, the question feels urgent again.
mindtheminds.substack.com/p/understand...
Understanding Without Thinking: John Searle, 1932-2025
The philosopher whose Chinese Room thought experiment still haunts AI researchers
mindtheminds.substack.com
October 6, 2025 at 9:40 AM
Philosopher John Searle died last month at 93. His Chinese Room Argument—can computers ever truly understand, or just manipulate symbols?—defined the AI consciousness debate.
With ChatGPT & Claude everywhere, the question feels urgent again.
mindtheminds.substack.com/p/understand...
With ChatGPT & Claude everywhere, the question feels urgent again.
mindtheminds.substack.com/p/understand...
🔗 Just enriched 1,000 records with demographic data using OpenRefine's reconciliation services in MINUTES
While everyone's reaching for LLMs, sometimes the best solution is a tool that's been perfecting one job for 10+ years
60% coverage via Wikidata, zero API costs ✨
#DataScience #OpenRefine
While everyone's reaching for LLMs, sometimes the best solution is a tool that's been perfecting one job for 10+ years
60% coverage via Wikidata, zero API costs ✨
#DataScience #OpenRefine
September 26, 2025 at 9:07 AM
🔗 Just enriched 1,000 records with demographic data using OpenRefine's reconciliation services in MINUTES
While everyone's reaching for LLMs, sometimes the best solution is a tool that's been perfecting one job for 10+ years
60% coverage via Wikidata, zero API costs ✨
#DataScience #OpenRefine
While everyone's reaching for LLMs, sometimes the best solution is a tool that's been perfecting one job for 10+ years
60% coverage via Wikidata, zero API costs ✨
#DataScience #OpenRefine
New blog series: beginner → advanced data science practitioner 🚀
Not ML algorithms - the engineering practices & tooling that matter in production.
First post: IDE selection. More on testing, packaging, deployment coming.
#DataScience #Python
Not ML algorithms - the engineering practices & tooling that matter in production.
First post: IDE selection. More on testing, packaging, deployment coming.
#DataScience #Python
The Data Science Engineering Gap: Part 0 – Your Development Environment - Crow Intelligence
This is the first post in a series about bridging the gap from beginner programmer to advanced data science practitioner. This transition isn't just about learning more Python - it's about adopting th...
crowintelligence.org
September 25, 2025 at 3:18 PM
New blog series: beginner → advanced data science practitioner 🚀
Not ML algorithms - the engineering practices & tooling that matter in production.
First post: IDE selection. More on testing, packaging, deployment coming.
#DataScience #Python
Not ML algorithms - the engineering practices & tooling that matter in production.
First post: IDE selection. More on testing, packaging, deployment coming.
#DataScience #Python
Humans can refuse the Doomsday Machine. Machines can’t.
Why that matters for AI, nuclear strategy, and the toxin we can’t drink: mindtheminds.substack.com/p/intoxicate...
Why that matters for AI, nuclear strategy, and the toxin we can’t drink: mindtheminds.substack.com/p/intoxicate...
Intoxicate Me Now
AI, Nuclear Weapons, and the Paradox of Intention
mindtheminds.substack.com
September 24, 2025 at 12:31 PM
Humans can refuse the Doomsday Machine. Machines can’t.
Why that matters for AI, nuclear strategy, and the toxin we can’t drink: mindtheminds.substack.com/p/intoxicate...
Why that matters for AI, nuclear strategy, and the toxin we can’t drink: mindtheminds.substack.com/p/intoxicate...
"Privatisation" appeared in 0.0000012% of published words in 1940. By the 1980s: 0.0008%—a 670-fold increase.
Tracking 200 years of text shows how Reagan-Thatcher economics didn't just change policy—it rewired our language and thinking.
mindtheminds.substack.com/p/what-money...
Tracking 200 years of text shows how Reagan-Thatcher economics didn't just change policy—it rewired our language and thinking.
mindtheminds.substack.com/p/what-money...
What Money Already Bought
How Economic Narratives Rewired Our Language
mindtheminds.substack.com
September 23, 2025 at 12:03 PM
"Privatisation" appeared in 0.0000012% of published words in 1940. By the 1980s: 0.0008%—a 670-fold increase.
Tracking 200 years of text shows how Reagan-Thatcher economics didn't just change policy—it rewired our language and thinking.
mindtheminds.substack.com/p/what-money...
Tracking 200 years of text shows how Reagan-Thatcher economics didn't just change policy—it rewired our language and thinking.
mindtheminds.substack.com/p/what-money...
Jevons' paradox strikes again: as AI becomes more efficient, total energy consumption explodes. We're not heading for reduced power usage—we're building toward ubiquitous AI that needs infrastructure everywhere.
The efficiency trap that enables ubiquity.
mindtheminds.substack.com/p/following-...
The efficiency trap that enables ubiquity.
mindtheminds.substack.com/p/following-...
Following the Power
How AI's energy constraints are driving the infrastructure investments that will enable ubiquitous computing
mindtheminds.substack.com
September 23, 2025 at 12:01 PM
Jevons' paradox strikes again: as AI becomes more efficient, total energy consumption explodes. We're not heading for reduced power usage—we're building toward ubiquitous AI that needs infrastructure everywhere.
The efficiency trap that enables ubiquity.
mindtheminds.substack.com/p/following-...
The efficiency trap that enables ubiquity.
mindtheminds.substack.com/p/following-...
Why does it feel wrong when Sweden's PM consults ChatGPT for policy advice? Beyond security risks: AI lacks the intellectual virtues—honesty, humility, responsibility—that leadership requires.
Can't outsource judgment. mindtheminds.substack.com/p/vices-and-...
Can't outsource judgment. mindtheminds.substack.com/p/vices-and-...
Vices and Virtues of AI
Why Vibe Coding Policy Questions Is a Bad Idea (Just Ask the Swedish PM)
mindtheminds.substack.com
September 23, 2025 at 11:59 AM
Why does it feel wrong when Sweden's PM consults ChatGPT for policy advice? Beyond security risks: AI lacks the intellectual virtues—honesty, humility, responsibility—that leadership requires.
Can't outsource judgment. mindtheminds.substack.com/p/vices-and-...
Can't outsource judgment. mindtheminds.substack.com/p/vices-and-...
Recreated Kindleberger's haunting "spiral" chart showing global trade collapsing like water down a drain (1929-1933).
His "leadership gap" theory feels uncomfortably relevant today.
Built in Python. History's warning: the spiral doesn't lie.
substack.com/@zoltanvarju...
His "leadership gap" theory feels uncomfortably relevant today.
Built in Python. History's warning: the spiral doesn't lie.
substack.com/@zoltanvarju...
Zoltan Varju (@zoltanvarju)
Just finished recreating one of economics' most haunting visualizations—Charles Kindleberger's famous "spiral" chart that shows global trade collapsing like "water circling a drain" between 1929-1933....
substack.com
September 23, 2025 at 8:58 AM
Recreated Kindleberger's haunting "spiral" chart showing global trade collapsing like water down a drain (1929-1933).
His "leadership gap" theory feels uncomfortably relevant today.
Built in Python. History's warning: the spiral doesn't lie.
substack.com/@zoltanvarju...
His "leadership gap" theory feels uncomfortably relevant today.
Built in Python. History's warning: the spiral doesn't lie.
substack.com/@zoltanvarju...
From childhood memories of Cold War missiles to today's AI arms race - exploring how nuclear deterrence paradoxes might become terrifyingly literal when machines make split-second decisions
Can AI solve commitment problems or just eliminate human choice entirely? substack.com/@zoltanvarju...
Can AI solve commitment problems or just eliminate human choice entirely? substack.com/@zoltanvarju...
Zoltan Varju (@zoltanvarju)
One of my earliest memories: 1987, age six, asking my father what nuclear weapons were after hearing about Pershing missiles being withdrawn from Germany. He explained, then assured me the Soviets had...
substack.com
September 23, 2025 at 8:55 AM
From childhood memories of Cold War missiles to today's AI arms race - exploring how nuclear deterrence paradoxes might become terrifyingly literal when machines make split-second decisions
Can AI solve commitment problems or just eliminate human choice entirely? substack.com/@zoltanvarju...
Can AI solve commitment problems or just eliminate human choice entirely? substack.com/@zoltanvarju...
One month with Claude Code: makes obvious mistakes, questionable choices, but genuinely speeds up development like having a smart intern.
Tried PyCharm plugin + Zed extension, landed on Zed + CLI. Not perfect, but hits that "flawed but useful" sweet spot.
Tried PyCharm plugin + Zed extension, landed on Zed + CLI. Not perfect, but hits that "flawed but useful" sweet spot.
September 23, 2025 at 8:52 AM
One month with Claude Code: makes obvious mistakes, questionable choices, but genuinely speeds up development like having a smart intern.
Tried PyCharm plugin + Zed extension, landed on Zed + CLI. Not perfect, but hits that "flawed but useful" sweet spot.
Tried PyCharm plugin + Zed extension, landed on Zed + CLI. Not perfect, but hits that "flawed but useful" sweet spot.
Reposted by Zoltan Varju
Compositionality is a central desideratum for intelligent systems...but it's a fuzzy concept and difficult to quantify. In this blog post, lab member @ericelmoznino.bsky.social outlines ideas toward formalizing it & surveys recent work. A must-read for interested researchers in AI and Neuro
Very excited to release a new blog post that formalizes what it means for data to be compositional, and shows how compositionality can exist at multiple scales. Early days, but I think there may be significant implications for AI. Check it out! ericelmoznino.github.io/blog/2025/08...
Defining and quantifying compositional structure
What is compositionality? For those of us working in AI or cognitive neuroscience this question can appear easy at first, but becomes increasingly perplexing the more we think about it. We aren’t shor...
ericelmoznino.github.io
August 19, 2025 at 1:51 PM
Compositionality is a central desideratum for intelligent systems...but it's a fuzzy concept and difficult to quantify. In this blog post, lab member @ericelmoznino.bsky.social outlines ideas toward formalizing it & surveys recent work. A must-read for interested researchers in AI and Neuro
Reposted by Zoltan Varju
www.programmablemutter.com/p/large-lang... Gopnikism, interactionism, structuralism and role play.
Large language models are cultural technologies. What might that mean?
Four different perspectives
www.programmablemutter.com
August 18, 2025 at 1:00 PM
www.programmablemutter.com/p/large-lang... Gopnikism, interactionism, structuralism and role play.
Currently learning Canva to create better data visualizations for my blog. Working on mapping energy infrastructure and data center locations for an upcoming post on AI's resource demands.
#DataVisualization #Canva
#DataVisualization #Canva
August 15, 2025 at 2:04 PM
Currently learning Canva to create better data visualizations for my blog. Working on mapping energy infrastructure and data center locations for an upcoming post on AI's resource demands.
#DataVisualization #Canva
#DataVisualization #Canva
46% of Chinese firms use generative AI.
9% see real benefits.
The other 37%? At least their PowerPoints look amazing 😄
#AI #GenAI #Productivity
9% see real benefits.
The other 37%? At least their PowerPoints look amazing 😄
#AI #GenAI #Productivity
August 14, 2025 at 10:58 AM
46% of Chinese firms use generative AI.
9% see real benefits.
The other 37%? At least their PowerPoints look amazing 😄
#AI #GenAI #Productivity
9% see real benefits.
The other 37%? At least their PowerPoints look amazing 😄
#AI #GenAI #Productivity
In June last year, only 8% of Greater China firms surveyed by Gartner were using generative AI. Less than a year later: 43%. (The Economist)
Meanwhile, here in Hungary we’d be thrilled if even 8% of firms used AI 😄
#AI #GenerativeAI #DigitalTransformation
Meanwhile, here in Hungary we’d be thrilled if even 8% of firms used AI 😄
#AI #GenerativeAI #DigitalTransformation
August 14, 2025 at 10:08 AM
In June last year, only 8% of Greater China firms surveyed by Gartner were using generative AI. Less than a year later: 43%. (The Economist)
Meanwhile, here in Hungary we’d be thrilled if even 8% of firms used AI 😄
#AI #GenerativeAI #DigitalTransformation
Meanwhile, here in Hungary we’d be thrilled if even 8% of firms used AI 😄
#AI #GenerativeAI #DigitalTransformation
Reposted by Zoltan Varju
Some hope that generative AI will make starting a company so cheap and hassle-free that anyone could do it. Whether these entrepreneurs will be able to escape the grip of America’s tech giants is another matter
How AI could create the first one-person unicorn
The technology is allowing entrepreneurs to start and grow businesses on their own
econ.st
August 13, 2025 at 12:20 PM
Some hope that generative AI will make starting a company so cheap and hassle-free that anyone could do it. Whether these entrepreneurs will be able to escape the grip of America’s tech giants is another matter
📈➡️📉 Recreated Kindleberger's famous trade collapse "spiral" and turned it into an infographic with Canva.
The "leadership gap" concept feels especially relevant today...
Full story: crowintelligence.org/2025/05/12/v...
#DataViz #Economics #Kindleberger
The "leadership gap" concept feels especially relevant today...
Full story: crowintelligence.org/2025/05/12/v...
#DataViz #Economics #Kindleberger
August 13, 2025 at 12:15 PM
📈➡️📉 Recreated Kindleberger's famous trade collapse "spiral" and turned it into an infographic with Canva.
The "leadership gap" concept feels especially relevant today...
Full story: crowintelligence.org/2025/05/12/v...
#DataViz #Economics #Kindleberger
The "leadership gap" concept feels especially relevant today...
Full story: crowintelligence.org/2025/05/12/v...
#DataViz #Economics #Kindleberger
Quanta piece on law of increasing functional information - complexity grows everywhere: minerals to life to intelligence.
Echoes Chalmers: information has physical and phenomenal aspects - conscious experience is fundamental, alongside mass, charge, space-time
www.quantamagazine.org/why-everythi...
Echoes Chalmers: information has physical and phenomenal aspects - conscious experience is fundamental, alongside mass, charge, space-time
www.quantamagazine.org/why-everythi...
Why Everything in the Universe Turns More Complex | Quanta Magazine
A new suggestion that complexity increases over time, not just in living organisms but in the nonliving world, promises to rewrite notions of time and evolution.
www.quantamagazine.org
August 4, 2025 at 9:48 AM
Quanta piece on law of increasing functional information - complexity grows everywhere: minerals to life to intelligence.
Echoes Chalmers: information has physical and phenomenal aspects - conscious experience is fundamental, alongside mass, charge, space-time
www.quantamagazine.org/why-everythi...
Echoes Chalmers: information has physical and phenomenal aspects - conscious experience is fundamental, alongside mass, charge, space-time
www.quantamagazine.org/why-everythi...
I live COLD War era movies. I think they can teach us a lot about AI governance too crowintelligence.org/2025/06/16/w...
August 4, 2025 at 9:28 AM
I live COLD War era movies. I think they can teach us a lot about AI governance too crowintelligence.org/2025/06/16/w...