iiinova — where your digital self keeps evolving

Reposted by Fernando Cucchietti

ren faire @lathrys.at · 03/10/2026

the problem with ai for the kind of person with my kind of brain is not that it offloads hard cognitive work, per se, but rather it enables a certain disposition to run 24/7 best summarized as:

330239

Fernando Cucchietti @fercook.bsky.social · 03/10/2026

Absolutely not to your point, but I would prefer if they throw in calculators, word counters, etc to help increase the precision of answers.

000

Reposted by Fernando Cucchietti

Georgios Karamanis @karaman.is · 02/10/2026

I made a macOS screensaver that draws the warming stripes one year at a time. It holds all 1,069 locations from showyourstripes.info and nine ways to build each picture. Free download. Stripes by Ed Hawkins @edhawkins.org karaman.is/blog/2026/10... #ShowYourStripes #dataviz #macOS

A MacBook on a table outdoors, with autumn trees behind it. The screen fills with the warming stripes for Sweden, 1850–2025: thin vertical bars in blues and pale shades, turning to deep reds toward the right. A small label in the corner reads "Sweden 1850–2025"

26115

Reposted by Fernando Cucchietti

Kelsey Hightower @kelseyhightower.com · 02/10/2026

People really out here choosing airlines based on in-flight WiFi speeds.

15533

Fernando Cucchietti @fercook.bsky.social · 02/10/2026

You get cooties just by looking at the logo!

010

Reposted by Fernando Cucchietti

Zach Weinersmith @zachweinersmith.bsky.social · 30/09/2026

I think it's not a great argument against AI eating a lot of the arts to say that AI can't be creative/original. Setting aside the question of whether they can... does *most* art strike you as extremely original? The vast majority of people, including great artists most of the time are mid.

3723216

Fernando Cucchietti @fercook.bsky.social · 30/09/2026

Maybe it was like that when stakes are low...but still, ime about half where super competitive and secretive.

google.com

Google Search

001

Fernando Cucchietti @fercook.bsky.social · 29/09/2026

This is an awesome idea for an obsidian plugin that takes meeting notes

000

Fernando Cucchietti @fercook.bsky.social · 28/09/2026

If there are any nonbelievers — also I want to brag

010

Fernando Cucchietti @fercook.bsky.social · 28/09/2026

I did butter chicken today! And I have leftovers now!

010

Fernando Cucchietti @fercook.bsky.social · 28/09/2026

A side issue on this debate is that frontier labs are recklessly posting things on blogs--not doing science, just pure marketing. Scientists take a long time between discovery and publishing because they need to make sure they are not pumping out bullshit. www.nytimes.com/2026/09/27/s...

nytimes.com

Did Anthropic’s A.I. Really Make a Scientific Discovery on Its Own?

An expert at the University of Copenhagen said his team had been sharing its research with the company’s A.I. model, Claude, and that its new finding matched their work.

000

Fernando Cucchietti @fercook.bsky.social · 28/09/2026

Looks like AI has the same urge I have when solving puzzles

Drawing of four puzzle pieces that are clearly not meant to be close but forcibly put together

000

Reposted by Fernando Cucchietti

Matt Henderson @matthen.com · 26/09/2026

How have the constellations changed since humans first looked up? Has the big dipper always pointed to Polaris? Has Polaris always been the North star? (nope!)

211119

Fernando Cucchietti @fercook.bsky.social · 25/09/2026

The Claude code moment for robots is happening now

020

Reposted by Fernando Cucchietti

David Aerne @meodai.bsky.social · 17/09/2026

CuspHanger 0.4.0 gives you control over the lightness steps, in the form of a bezier curve or any other function. meodai.github.io/cusphanger/

25612

Fernando Cucchietti @fercook.bsky.social · 25/09/2026

Two things: 1. Just lovely I find these kids so cute and creative. I say that if they manage to hack the system they deserve the rewards. 2. I just realized (again) I am an idiot, somehow my brain mixed Ezra Klein with Ira Glass, believing now he was at NYT he had gotten so bad.

010

Reposted by Fernando Cucchietti

Aaron Becker @aaronjbecker.bsky.social · 22/09/2026

You could find your way along the Nova Scotia coast blind, just by timing the blinks. Every lighthouse flashes on its own rhythm, a set interval and colour no neighbour repeats. Sambro Island off Halifax has burned since 1758, the oldest working lighthouse in North America. #lighthouse #NovaScotia

815157

Fernando Cucchietti @fercook.bsky.social · 24/09/2026

Because people famously eat concentrated whey, just like a chicken breast

031

Fernando Cucchietti @fercook.bsky.social · 23/09/2026

Also price, a task I was running for €1.60 a pop now costs €1.10

010

Fernando Cucchietti @fercook.bsky.social · 22/09/2026

Wait it would be super general intelligence right? SGI ftw

100

Fernando Cucchietti @fercook.bsky.social · 22/09/2026

Who will pay for this Fable then?

010

Fernando Cucchietti @fercook.bsky.social · 22/09/2026

Amodei letter: everyone should pause to give time for our biolabs to churn out dogs OpenAI: announces creation of dog council to opine on the dog creation efforts Grok: we can draw puppies Google: come on now dogs?

000

Fernando Cucchietti @fercook.bsky.social · 22/09/2026

New Siri doesn't get 30% of what I say compared to previous one 🤷

010

Reposted by Fernando Cucchietti

K @kerry.bsky.social · 19/09/2026

“I Built Non-Autoregressive Decision Models with RL a Year Ago. Then a Frontier Lab Called It a "Breakthrough".” Before Jev there was Laya laya.convaiinnovations.com (via @dherman.dev)

laya.convaiinnovations.com

Laya — 33ms Multilingual System 1 Decision Engine

Evaluates typed decisions (choice, score, noul) over 100+ languages in a single forward pass with calibrated probabilities. Outperforms TypeSafe Jev.

15924

Fernando Cucchietti @fercook.bsky.social · 19/09/2026

Now, more than ever, the important thing is knowing what to do

001

Fernando Cucchietti @fercook.bsky.social · 19/09/2026

I think we already proved that model with the millions of Nigerian prince inheritance emails costing almost nothing.

100

Fernando Cucchietti @fercook.bsky.social · 19/09/2026

Most carmakers are making terrible UIs and underestimating its importance and how troyan horse CarPlay-like tech will eat a large % of their market

020

Fernando Cucchietti @fercook.bsky.social · 19/09/2026

A related measurement is when the annoyance of working on it is larger than the potential embarrassment of something not being correct

020

Reposted by Fernando Cucchietti

Jer Thorp @jerthorp.bsky.social · 18/09/2026

Friends! Can I ask you a favour? My 🐦 + 📊 course starts next week. I'm super excited but registration is worryingly low. Can you share this with your weirdo nature + art + code friends? jerthorp.me/learning

jerthorp.me

LEARNING | jerthorp

Online courses and wokshops about birds, data and creative coding. Binoculars to Binomials is for coders who are interested in cultivating an observational practice, and for birders who want to dive ...

25357

Fernando Cucchietti @fercook.bsky.social · 17/09/2026

Solopreneur? Is that what they call it?

020

Fernando Cucchietti @fercook.bsky.social · 15/09/2026

Wait don’t they say they post-trained K2.6?

100

Fernando Cucchietti @fercook.bsky.social · 14/09/2026

Long project development with AIs can be tiresome

010

Fernando Cucchietti @fercook.bsky.social · 14/09/2026

1. Well, they do have some of the most expensive models. 2. Just imagine the profits if they could, I don't know, slow or pause development a bit and spend less on training

030

Fernando Cucchietti @fercook.bsky.social · 14/09/2026

oh boy, I did not know that was even possible (unless CEO-types)

000

Fernando Cucchietti @fercook.bsky.social · 13/09/2026

No se Javier, suena a marketing.

000

Reposted by Fernando Cucchietti

Ethan Mollick @emollick.bsky.social · 13/09/2026

Those effects will not be all at once, but the fact that change will occur is inevitable now. But inevitable change does not mean the type of change is inevitable. It is increasingly important that we develop & share models of AI-human work that enhance, rather than only replace, human labor.

1995

Fernando Cucchietti @fercook.bsky.social · 13/09/2026

Nine figures? As in more than 100 million and less than a billion per year? In salary?

2100

Fernando Cucchietti @fercook.bsky.social · 13/09/2026

Node based programming, or just connecting things ala mind maps or miro would be great with this interface

000

Fernando Cucchietti @fercook.bsky.social · 13/09/2026

There is no “our server” when you have 100.000s of nodes and all tasks are at the TB scale. Even detecting the traffic is hard, Eg Cloudflare moves 500 Tbps around the world

000

Reposted by Fernando Cucchietti

Carlos A. Scolari @cscolari.bsky.social · 12/09/2026

La idea del fin del mundo tiene cara de algoritmo, avanza a paso de ganso y suda la gota gorda del cambio climático. Pero,¿hubo alguna vez un momento donde la sensación de fin se expresara simultáneamente en el plano tecnológico, político y climático? 🔥🔥🔥 hipermediaciones.com/2026/09/05/e...

hipermediaciones.com

El volcán del fin del mundo.

Estamos en un momento de la historia donde se extiende la sensación de que el mundo en que vivíamos -con sus urgencias, obsesiones, esperanzas y ritmos- se derrite como un alfajor Havanna de chocol…

035

Reposted by Fernando Cucchietti

Simon Willison @simonwillison.net · 09/09/2026

Wrote up my thoughts on the whole OpenAI Navier–Stokes Millennium Prize Problem story, and how it highlights the still confusing question of what using my data "to improve model performance" actually means simonwillison.net/2026/Sep/8/o...

This also highlights one of my ongoing frustrations about how all of this works. When an AI lab says that my data is "used to improve model performance", what does that actually mean?

My two favourite hypothetical questions regarding this used to be:

If I'm running Codex and one of my API keys accidentally gets consumed in the context, what are the chances that someone else might ask for an API key in the future and get mine back? (I asked someone at OpenAI once and they called this the "regurgitation" problem and assured me that they take great pains to prevent that... but wouldn't describe how.)
If I brainstorm with ChatGPT about potential new directions for my company, what's the chance that information might be exposed to a competitor in six months' time who asks "what might company X plan to do next"?

My new preferred hypothetical for this is:

If I use ChatGPT to help me partially solve a Millennium Prize problem, what are the chances that my work will influence training such that a later model helps someone else solve it first?

1129052

Reposted by Fernando Cucchietti

Ted Underwood @tedunderwood.com · 08/09/2026

So what used to be the case: it’s almost always folly to worry about being scooped, because there’s no way mere knowledge of your direction of inquiry could allow a competitor to reproduce all your grunt work quickly enough. That’s the advice I *used* to give.

2887

Fernando Cucchietti @fercook.bsky.social · 08/09/2026

What I don’t get is why isn’t this a scandal for companies that trust OpenAI with sensitive information

0100

Fernando Cucchietti @fercook.bsky.social · 06/09/2026

The graphic design is also…interesting

010