Lightnews — Scholar-powered news

nzagalo @nzagalo.bsky.social · 2d

The Grand AGI Delusion by @garymarcus.bsky.social

The Grand AGI Delusion

and a fantastic symposium on AI and society at The Royal Society

open.substack.com

nzagalo @nzagalo.bsky.social · 4d

As organizations integrate generative AI into their daily routines, they face a growing paradox: increased adoption coinciding with diminishing trust.

Professional Disenchantment of AI

When Error Overthrows Functional Belief

open.substack.com

1

Reposted by nzagalo

Miguel, o fleumático 🚶🏻 @tomarnogueira.bsky.social · 5d

Já escrevi muitos posts a sublinhar o óbvio. Já alertei muita gente à minha volta para a falácia e acabo sempre a ser olhado como o velho do Restelo a gritar para os barcos. O @nzagalo.bsky.social escreve aqui o melhor resumo de todos. Ele só não acha que vai haver um colapso. Eu acho que vai.

The OpenAI House of Cards

From curing the brain to selling the feed, the short life of a modern miracle.

mirrorsofthought.substack.com

1 1 1

nzagalo @nzagalo.bsky.social · 5d

Eu não falo no colapso porque não sou especialista em economia. Tenho aí um rascunho sobre isso, as condições perfeitas estão no ar. Mas a OpenAI está a tentar "tornar-se grande demais para colapsar". Vamos ver.

1 1

Reposted by nzagalo

Ian Fraser @ianfraser.bsky.social · 9d

Dr Christopher Rudge, law lecturer at the University of Sydney, says the report that Deloitte prepared for Australia's Department of Employment and Workplace Relations contained “hallucinations” because of an initially undisclosed use of generative AI. Now the government is getting its money back.

Deloitte to pay money back to Albanese government after using AI in $440,000 report

Partial refund to be issued after several errors were found in a report into a department’s compliance framework

www.theguardian.com

37 40

nzagalo @nzagalo.bsky.social · 9d

"the broader labor market has not experienced a discernible disruption since ChatGPT’s release 33 months ago, undercutting fears that AI automation is currently eroding the demand for cognitive labor across the economy."

-- Yale University Report, October 1, 2025

Evaluating the Impact of AI on the Labor Market: Current State of Affairs

budgetlab.yale.edu

2

nzagalo @nzagalo.bsky.social · 10d

US = China

ICE Wants to Build Out a 24/7 Social Media Surveillance Team

Documents show that ICE plans to hire dozens of contractors to scan X, Facebook, TikTok, and other platforms to target people for deportation.

www.wired.com

1

nzagalo @nzagalo.bsky.social · 12d

OpenAI in 2021: “we want to cure brain cancer”

OpenAI in 2024: "we need far more compute, and therefore far more capital -- in order to pursue our mission."

OpenAI in 2025: Launch a new new Social App.

1

nzagalo @nzagalo.bsky.social · 14d

"while models beat humans on benchmarks, the standardized tests designed to measure AI performance, they struggle to replicate this performance in hospital conditions."

AI isn't replacing radiologists

Radiology combines digital images, clear benchmarks, and repeatable tasks. But demand for human radiologists is at an all-time high.

www.understandingai.org

1 2

nzagalo @nzagalo.bsky.social · 15d

A arena das redes sociais foi tomada pelo Chega, é preciso entrar no ringue. Não se trata de copiar a raiva populista. Trata-se de reinventar a comunicação com emocionalidade positiva. A linguagem institucional pode coexistir com formatos digitais participativos.

Xadrez contra Wrestling: Política Tradicional vs. Trump/Chega

Nos últimos dois anos fui entrevistado por vários órgãos de comunicação social, onde avisei que a arena das redes sociais estava…

medium.com

nzagalo @nzagalo.bsky.social · 15d

"Yet, there is little research that explains how educators actually decide when, how, and why they choose whether or not to use AI tools."

How Math Teachers Are Making Decisions About Using AI | Stanford HAI

A Stanford summit explored how K-12 educators are selecting, adapting, and critiquing AI tools for effective learning.

hai.stanford.edu

1

nzagalo @nzagalo.bsky.social · 15d

Your instance of ChatGPT helped you clarify some ideas on a thorny problem (perhaps related to AI itself..) that you'd been thinking about for ages, but had never quite managed to get over that last hump. Now, however, with its help (and encouragement), you've arrived at truly profound conclusions.

So You Think You've Awoken ChatGPT — LessWrong

Written in an attempt to fulfill @Raemon's request. …

www.lesswrong.com

1 2

nzagalo @nzagalo.bsky.social · 15d

"tadpoles that switch sex are “making it to adulthood, becoming sexually viable, and reproducing successfully.""

The Queer Lives of Frogs

What frogs teach us about sex, science, and why biology is messier than we think.

thereader.mitpress.mit.edu

nzagalo @nzagalo.bsky.social · 18d

Feito. Mas quase desisti na C4.

4

nzagalo @nzagalo.bsky.social · 18d

😂

1

nzagalo @nzagalo.bsky.social · 18d

"Violence against the Truth"

Barack Obama @barackobama.bsky.social · 19d

We have people in power making broad claims around certain drugs and autism that have been continuously disproven.

1

nzagalo @nzagalo.bsky.social · 18d

"Practically the only people still pretending scaling LLMs is “all you need” are grifters."

Game over for pure LLMs. Even Turing Award Winner Rich Sutton has gotten off the bus.

One by one, all the big names have turned around. What should we do next?

garymarcus.substack.com

2 4

nzagalo @nzagalo.bsky.social · 18d

"There is no productivity gain if someone wastes hours searching for the “perfect” prompt for something the model cannot essentially do."

Why Prompt Engineering Should Not Be Taken Seriously

Why there are no correct and incorrect prompts

msukhareva.substack.com

1

nzagalo @nzagalo.bsky.social · 18d

"In collaboration with Stanford Social Media Lab, our research team at BetterUp Labs has identified one possible reason: Employees are using AI tools to create low-effort, passable looking work that ends up creating more work for their coworkers."

AI-Generated “Workslop” Is Destroying Productivity

Despite a surge in generative AI use across workplaces, most companies are seeing little measurable ROI. One possible reason is because AI tools are being used to produce “workslop”—content that appea...

hbr.org

1 1

nzagalo @nzagalo.bsky.social · 20d

“no amount of prompt engineering guarantees the right answer”

Why Prompt Engineering Should Not Be Taken Seriously

Why there are no correct and incorrect prompts

msukhareva.substack.com

2

nzagalo @nzagalo.bsky.social · 20d

There is a huge gap between investment and revenue from LLMs. Investors wrongly assume everyone will be a winner.

AI investors are in for a rude awakening | Roger McNamee

There is a huge gap between investment and revenue from LLMs. Investors wrongly assume everyone will be a winner

www.theguardian.com

1

nzagalo @nzagalo.bsky.social · 23d

Ainda não sabemos como vai ser. É um bom espaço para inovar. A tecnologia está aí, é necessário explorar, experimentar, criar :)

1

nzagalo @nzagalo.bsky.social · Sep 4

“Suddenly, I was watching him use ChatGPT,” says Declan, 31, who lives in Los Angeles. “He was taking what I was saying and putting it into ChatGPT, and then summarizing or cherry-picking answers.”

Therapists are secretly using ChatGPT. Clients are triggered.

Some therapists are using AI during therapy sessions. They’re risking their clients’ trust and privacy in the process.

www.technologyreview.com

1 2

nzagalo @nzagalo.bsky.social · Sep 3

My new book Artificial Media (Springer) is out. But beyond announcing, I ask: what does it mean to create with machines? In Mirrors of Thought I explain why artificial media is not just a trend, but a paradigm shift in creativity and culture.

open.substack.com/pub/mirrorso...

2 7

nzagalo @nzagalo.bsky.social · Sep 3

😂