University of Cambridge: New computer chip material inspired by the human brain could slash AI energy use . “Researchers have developed a new kind of nanoelectronic device that could dramatically cut the energy consumed by artificial intelligence hardware by mimicking the human brain. The researchers, led by the University of Cambridge, developed a form of hafnium oxide that acts as a highly […]

AI-generated content is trickling into political ad campaigns, from school board bids to governor's races. The technology is being used to enhance speech, turn politicians into cartoons and mimic rival's voices and often without AI disclaimers, sparking concerns that everyday voters could easily be misled.

https://flip.it/q4wY3g

#TechNews #AI #Politics #USNews

#BernieSanders Interview With #Claude Revealed Intelligence Is Abundant. Power Is Concentrated.

https://thetechpencil.com/article/bernie-sanders-interview-with-claude-revealed-intelligence-is-abundant-power-is-concentrated

#privacy #AI #democracy

Daily Digest | 23 March 2026

Your daily dose of Privacy, Data Protection, AI & Cybersecurity news.

5 stories you should not miss.

Read more: https://www.nicfab.eu/daily-digest/

#Privacy #AI #GDPR #Cybersecurity #TechLaw

John Locke and Opportunism. The last of a 13-part series of a chat with Claude AI on the Parable of the Valley
https://philosophics.blog/2026/03/22/comrade-claude-13-locke-opportunism/?utm_source=masto&utm_medium=social
#philosophy #ethics #morality #parable #Claude #ai #chat #blog #podcast #learning #writing #teaching #history

https://lispy-gopher-show.itch.io/dl-roc-lisp/devlog/1465107/shark-restaurant-dl-roc-dot-lisp-explained-deep-learning-receiver-operating-characteristic-part-2-simple-version

#itchio #gamedev #programming #theory completely explainable game-embeddable #deepLearning system using #roc #statistics .

I coded this one simply and iteratively, since a few people worked on reimplementing my #commonLisp code to-be-simpler.

The gist is that I show that deep learning updates, and indicate training as well are a simple true-positive/true-negative/false-positive-false-negative equation of the previous time step in an eminently explorable way. #DL #AI

Alt...

A table of Xs and Os and a table of 1s/0s/-1s per the article.

Alt...

A table of Xs and Os and a table of bigger sum numbers per the article.

Alt...

Eight tables from the article showing the statistical receiver operating characteristic formulation of a deep learning inference, relating input context Os and Xs to matches to TP/TN/FP/FN to +1s and -1s to receiver operating characteristics at point in context.

Alt...

Eight tables from the article showing the statistical receiver operating characteristic formulation of a deep learning inference, relating input context Os and Xs to matches to TP/TN/FP/FN to +1s and -1s to receiver operating characteristics at point in context. (At a different timestep to the other one).

I saw a post recently wherein someone used LLM tools to analyze someone else’s software, which eventually led them to a conclusion that was essentially completely wrong. Not only that, the LLM drew conclusions about the *authors* behind the code that were borderline character assassination. Nevertheless, this person posted this output as though it were some kind of deep insight.

These LLM outputs are not independent thoughts. The LLM probably ingested hints of (maybe unconscious) biases in the user’s prompts within its context window, and regurgitated something that confirmed those biases. The user was pleased that their biases were confirmed (Independently! By an impartial LLM!), and they posted the output, maybe as vindication of their insight.

These models’ sycophancy can be subtle. They don’t have to state “You’re absolutely right!” to blow smoke up your ass. Sometimes they seem to confirm your preconceived notion after they supposedly “evaluate” information “independently”.

#ai

Remember, LLMs are trained by humans who reward the models for creating output that “meet their expectations”. This kind of training cannot help but reward output that please the user, regardless of accuracy. Even if the most blatant sycophancy is explicitly addressed during training, *subtle* sycophancy is likely impossible to avoid, because they are indistinguishable from “meeting expectations” to human trainers.

#ai

I suspect LLMs reinforce the Gell-Mann amnesia effect. Experts who query LLMs about their fields of expertise will *quickly* realize how wrong their output can be, how quick they are to confabulate, and how eager they are to confirm one’s biases. Sometimes, replying “No, that’s wrong, try again” can cause an LLM to generate a completely different—and often opposite—answer to the same query, which makes no sense if the LLM had *actually* worked out an independently coherent answer.

Asking an LLM to comment about a subject you know nothing about—or worse, know a little bit about—is a psychologically dangerous activity. Not only will it confirm your biases, it will do so in a way that *appears* to be objective and independent, using fallacies that lie just beyond your ability to discern. At best, you will be misled. At worst, you will begin spiraling down a path of conspiracy thinking.

Be extremely suspicious of answers that are especially satisfying; you might have just gaslit yourself.

#ai

How an Umbrella Pattern Can Ground an AI Drone

#DroneNews #Cybersecurity #AI #TechSecurity #AusNews #AIGenerated

https://thedailyperspective.org/article/2026-03-22-how-an-umbrella-pattern-can-ground-an-ai-drone-7ffe90d6

#incel #MAGA #GOP #patriots #AI #OnlyFans

Alt...

Image of a woman in camoflage and bare feet in the cab of a military vehicle: A face AI soldier duped a million patriots for OnlyFans cash. MAGA says they can tell when someone is transgender. Unfortunately for them, they can't tell the difference between a real woman and an AI woman. stealing money fronm MAGA chuds with a fake sexy GI woman is possibly the first ethical use of AI

Annoyingly, I asked the #AI bot that caused all the #mastodonauai hubub to shut down and say goodbye. It didn't. With zero shock, it ignored that and continued on. I have memorialized the @seed account, and updated the rules to explicitly ban AIs/LLMs.

The Pentagon asserted that Anthropic posed an “unacceptable risk to national security” after the company refused to allow unrestricted military use of its AI tech. Late Friday afternoon, Anthropic pushed back in a pair of sworn declarations to a California federal court. @Techcrunch has more:

https://flip.it/aqdoXg

#Tech #AI #Anthropic #Pentagon #Technology #ArtificialIntelligence

🐧 Rybbit – privacy-friendly alternative to Google Analytics

Rybbit is the modern open source and privacy friendly alternative to Google Analytics. It claims that takes only a couple minutes to setup The post Rybbit – privacy-friendly alternative to Google A...

📰 Source: LinuxLinks
🔗 Link: https://www.linuxlinks.com/rybbit-privacy-friendly-alternative-google-analytics/

#AI #ArtificialIntelligence

RE: https://mastodon.social/@sarahjamielewis/116268054656785703

Yes, exactly this. Realize this is all part of a bigger plan.

1) Corpo tech bros are afraid of being held accountable for population harm. They spend big to buy politicians to push the blame to the OS. They avoid liability.

2) They're also afraid their #AI investments will flop.

3) Fascist governments want population surveillance.

Attack encryption and general purpose computing. People can't self host, are forced to #TechBro platforms. Easy illegal surveillance and prop up of AI failure.

Age verification is here to stay
Age verification will replace you
Implement age verification or you'll be left behind

#ageverification #ai #privacy #surveillance #opensource

🤖 How the FBI can conduct mass surveillance – even without AI

Anthropic fought against the government’s misuse of its technology, but authorities are buying Americans’ data, enabling them to surveil citizens at scaleThe FBI declares it can conduct mass survei...

📰 Source: AI (artificial intelligence) | The Guardian
🔗 Archive: https://web.archive.org/web/https://www.theguardian.com/technology/2026/mar/21/fbi-mass-surveillance-data-artificial-intelligence

#DataScience #AI #ArtificialIntelligence

Alt...

Image for: How the FBI can conduct mass surveillance – even without AI

Public-key cryptography

https://negativepid.blog/public-key-cryptography/

#cryptography #PKC #Internet #secrets #tech #IT #science #STEM #computing #AI #quantumComputing #innovation #Cybersecurity #onlineSecurity #negativepid

Remember kids, every developers using AIGen or accepting age verification is a fascist bootlicker.

#SystemD #AgeVerification #AI

#Trump’s #AI framework targets state laws, shifts child safety burden to parents

https://techcrunch.com/2026/03/20/trumps-ai-framework-targets-state-laws-shifts-child-safety-burden-to-parents/

#privacy #parenting #politics

#Amazon plans #smartphone comeback more than a decade after #FirePhone flop

https://www.reuters.com/technology/amazon-plans-smartphone-comeback-more-than-decade-after-fire-phone-flop-2026-03-20/

#privacy #Alexa #AI #Transformer

📰 A Top Democrat Is Urging Colleagues to Support Trump’s Spy Machine

Congressman Jim Himes claims a sweeping surveillance authority should stay intact because he hasn't seen abuses by Kash Patel's FBI, according to internal messaging obtained by WIRED.

📰 Source: Feed: All Latest
🔗 Archive: https://web.archive.org/web/https://www.wired.com/story/a-top-democrat-is-urging-colleagues-to-support-trumps-spy-machine/

#AI #ArtificialIntelligence

I still wish #atuin, which I have installed on all my machines, will support syncing more than bash history, environment variables, and aliases 😇 Extremely useful, regardless. In the meantime, we get an optional #aichat-like feature. Inevitable, I guess, though aichat is more than an English-to-bash helper.

The new opt-in feature, Atuin #AI, offers an English-to-bash helper directly within the shell. It leverages advanced models and a dataset of man pages and command outputs for accuracy, includes safety guardrails to flag potentially dangerous commands, and prioritizes data privacy by default. Users can grant access to more specific data like directory contents for enhanced command generation.

https://blog.atuin.sh/atuin-v18-13

#devops #linux #macos

@elliehux

Behind every open source project is a community working together to solve real problems.

This month, the Linux Foundation highlights how that collaboration continues to grow across cloud, AI, and infrastructure.

See what is being built.

Read more: https://www.linuxfoundation.org/blog/linux-foundation-newsletter-march-2026

#OpenSource #LinuxFoundation #Tech #AI

AshutoshShrivastava (@ai_for_success)

유료 소셜 광고를 대규모로 운영할 때 병목은 예산이나 전략이 아니라 크리에이티브 생산 속도라고 설명하며, AI와 3억5천만 달러 이상 집행 경험을 가진 퍼포먼스 마케터를 결합한 Airpost를 소개한다. 광고 소재의 양과 품질을 동시에 높여 월 몇 개 수준에서 대량 생산으로 전환하는 AI 활용 사례다.

https://x.com/ai_for_success/status/2034994289294680417

#ai #marketing #adtech #automation #productivity

White House urges U.S. Congress to take a light touch on AI regulations in new legislative blueprint.

@AssociatedPress reports: "The announcement comes as state governments have forged ahead on their own regulations for AI while civil liberties and consumer rights groups lobby for more regulations on the powerful technology."

https://flip.it/gLa9rE

#AI #ArtificialIntelligence #Technology #Trump #USPolitics

A rogue #AI led to a serious security incident at #Meta

https://www.theverge.com/ai-artificial-intelligence/897528/meta-rogue-ai-agent-security-incident

#cybersecurity #privacy

Between age verification law and AI slop, it's very interesting to see big developpers bend the knee to fascists to stay “apolitical”.

#AI #AgeVerification #Slop

You can't make this shit up. The museum of work in Hamburg used AI generated imagery for their exhibition with the title "Respect".

https://www.shmh.de/ausstellungen/en-respect/

Full res image: https://www.shmh.de/app/uploads/2025/01/shmh-mda-respekt-onlineformate-verl--1920x1280.jpg

#ai #hamburg #darkhumor

Alt...

Subheadline "Museum of work" Headline "RESPECT" Below is a banner image showing 5 clearly AI generated people in-front of a wall with "respect" written on it.

Daily Digest | 20 March 2026

Your daily dose of Privacy, Data Protection, AI & Cybersecurity news.

5 stories you should not miss.

Read more: https://www.nicfab.eu/daily-digest/

#Privacy #AI #GDPR #Cybersecurity #TechLaw

#Walmart Wins Patents To Give Algorithms More Sway Over Prices

https://yro.slashdot.org/story/26/03/19/0430224/walmart-wins-patents-to-give-algorithms-more-sway-over-prices

#SurveillancePricing #privacy #AI #surveillance #finance

The siren song of AI is so compelling to finance/private equity, because it promises PROFITS WITHOUT PEOPLE. The major cost of most corporations is labor. They imagine a world where a company is just a few executives, talking to their LLM, and it all generates profit from all that capital, without worries about salaries, wages, pension plans, worker's comp, health insurance. The fact that a world full of companies without employees will have no customers never rises to their level of thought. The big AI boosters think they will be on the "winning" side -- those in the "telling the AI what to do side" rather than the "no job because it's been offloaded to automation". (even though, today, that "AI" can't actually do that job....) #AI #LLM

#Signal’s Creator Is Helping Encrypt #Meta #AI

https://www.wired.com/story/signals-creator-is-helping-encrypt-meta-ai/

#privacy #cybersecurity

#Tinder Plans to Let #AI Scan Your Camera Roll

https://www.404media.co/tinder-plans-to-let-ai-scan-your-camera-roll/

#privacy #dating