Luis Omar

Computing fan.

[Isaac Miller 🧩 @isaacbmiller1]

DSPy 3.4.0 was just released!

This release includes native support for Jev and System one models inside of DSPy! Use it with compatible signatures. This release also includes a brand new optimizer, ReAnchor, specifically for calibrating outputs with confidence.

[Ricard Solé @ricard_sole]

Can AI spread like a virus and erode cognitive autonomy? Our new paper models how LLM adoption can cross tipping points, triggering runaway dependence and tech lock-in, suggesting routes to "cognitive immunization". https://t.co/4e8ya6JNsW

[Luis Omar @ionekr]

@cgershen Take my money!

[Carlos Gershenson @cgershen]

This book – 15 years in the making – was possible thanks to so many people. It is certainly a collective effort. Finally, preorders are available (link below).

[Samantha Pérez Dávila @ysamanthap]

Yesterday I defended my PhD dissertation on cannabis markets in Mexico and California. Three main highlights I want to share:

[Eduardo G. Altmann @EduardoGAltmann]

Our paper on fractals in r-tipping has been published today as

Chaos 36, 083122 (2026)
https://t.co/0BLGYxSsKa

Many improvements over the pre-print, thanks to the very helpful comments by the editor and reviewers from Chaos @AIP_Publishing.

[Vovô CDO @san_alehandro]

@TELMEXSoluciona hola. Disculpa la molestia pero ya van dos semanas q mi servicio está offline y los técnicos solo no vienen. Aunque ya se reporto más de un vez

[Guillermo de Anda-Jáuregui @gdeandajauregui]

Y para festejar... https://t.co/SyLBPPFwkl

Share and enjoy! Pronto les platicaremos más al respecto...

[Ole Peters @ole_b_peters]

Trying something new. I let AI generate a podcast about this foundational ergodicity economics paper.
I've spotted only a few inaccuracies, but I think they're excusable at this level of presentation.

What do you think? More of this? Add to website?

Enjoy!

[Sean McClure @sean_a_mcclure]

For years I've explored ideas on my NonTrivial podcast.

I just launched a new channel that compresses the core messages in these episodes as short, focused explainer videos.

Learning, decision-making, careers, incentives, human behavior, complexity, and everyday life.

A shorter, consumable way to learn some of the patterns I've discovered over the years.

Here's one for the first NonTrivial episode I did.

Enjoy

Channel: https://t.co/yDRcg21Mgd

Video #1: https://t.co/yhOSpSDZP2

[DuckDuckGo @DuckDuckGo]

Day 2 of trying to find Merlin: the search continues!

Merlin, please DM us for an important business meeting. 🦆🤝🦆

[Luis Omar @ionekr]

@fantasmasuarez @FIFAcom A

[Aravind Srinivas @AravSrinivas]

Monitor the Situation
https://t.co/TLkTfvJLLR

[Yaneer Bar-Yam @yaneerbaryam]

The MV Hondius hantavirus outbreak isn't the first Andes virus outbreak. Attached article in 2020 documents person-to-person transmission in Argentina outbreak: 34 cases, 11 deaths, four generations of human-to-human transmission.
https://t.co/n262nS05l0

[Eduardo ⭔ @toorandom]

Recently I had the opportunity to talk about #AI and philosophy applied to #cybersecurity processes and automations using artificial intelligence. Here I share some slides. 1/6

[alex zhang @a1zhang]

New mini experiment + blogpost + trajectories!

tldr; we boost performance of RLM(GPT-5.2) to double the best performing number (38.7% --> 65.6%) on LongCoT-mini without any training! An example of the mismanaged geniuses hypothesis (MGH) we (@zli11010, @lateinteraction) proposed earlier this month.

The LongCoT benchmark showed that frontier LMs and RLMs struggled to solve difficult compositional reasoning tasks. The paper generally attributes this to the RLMs inability to perform task decomposition, but we argue this is more our fault in how we prompt them; this capability is fully available to GPT-5.2 with an RLM harness!

Building on @raw_works's insightful blogpost and @sumeetrm / @CharlieLondon02 et al.'s incredibly useful benchmark, where they originally found RLMs to be incapable of solving the MATH and CS splits altogether. We did not train anything since the release of the initial benchmark.

To be fully transparent, these results are not meant to be added to their leaderboard either; benchmarks measure isolated capabilities, and we focus on showing (through different, rather specific prompting) that the capabilities required to solve these tasks are available to the models without additional training! It also has implications about how we would go about training these systems. Full blog below, it's a nice read :)

[Kol Tregaskes @koltregaskes]

Ex-MIT researcher Isaak Freeman quits his PhD and drops the 50,000 H100 GPU roadmap to emulate a full human brain.

He mapped the entire path from 302-neuron worm to 86-billion-neuron human with connectomics costs now at 100 dollars per neuron and data acquisition via advanced microscopes as the only blocker left - digital humans just got a realistic timeline.
https://t.co/kGB5hOAQHC

[Luis Omar @ionekr]

@francoisfleuret I appreciate your kindness! I learnt many things, I guess

[François Fleuret @francoisfleuret]

Reminder that I wrote a little book about deep-learning, which is phone-formatted, entirely free, and nearing the 1M download:

https://t.co/7VnowYjJLY

[Mario Valle Reyes 🚩🚩🚩 @bilbeny]

Me pusieron a pensar el otro día y más allá de lo que estamos preparando con el taller de AI de @investor_camp que ha tenido mucho éxito desde el año pasado, haré la próx. semana un TALLER DE 2 DÍAS para aprender OpenClaw sin hype y de manera segura.
https://t.co/UIwPSS87ax