@Liberationtech profile picture
Liberation Technology @Liberationtech

One of the world's leading communities for tech for good. Founded at @Stanford in 2008. @nytimes featured. Helped launch @joindiaspora, now @WeVoyce.

Joined January 2011
  • Tweets 166,520
  • Following 33,068
  • Followers 39,154
2,797 Photos and videos
Google's DiffusionGemma generates 256 tokens in parallel and self-corrects as it goes venturebeat.com/technology/g… via @VentureBeat

Google's DiffusionGemma runs text 4x faster

Google's open-source diffusion language model generates 256 tokens in parallel and self-corrects, hitting 4x speed on one GPU at a cost to quality.

venturebeat.com
77
Big Tech, big cons: Scammers are hiding in the apps that make your life easy restofworld.org/2026/tech-fa…

Big Tech, big cons: Scammers are hiding in the apps that make your life easy

In her new book “Bharat Bluff: Inside the Cons of India’s Internet Revolution,” writer Soumya Gupta explores how fraudsters are exploiting the trust people place in big tech platforms like Google,...

restofworld.org
65
Hand-cranked AI box lets you get a workout while you wait for answers theregister.com/offbeat/2026…

Hand-cranked AI box lets you get a workout while you wait for answers

We're all familiar with AI cranks by now, but what about crank-powered AIs?

theregister.com
57
Reddit now lets users reply to any post with video comments thenextweb.com/news/reddit-v… via @thenextweb

Reddit now lets users reply to any post with video comments

Reddit launched video comments across all public SFW communities. Users can upload or record short-form video replies alongside text, images, and GIFs.

thenextweb.com
52
Alaskans will be flying blind after NSF decommissions ocean monitoring network arstechnica.com/science/2026…

Alaskans will be flying blind after NSF decommissions ocean monitoring network

Alaska's multibillion-dollar fishing industry and vulnerable coastal communities at risk.

arstechnica.com
54
Crackdown on tech platforms will go ahead despite US intervention, says No 10 theguardian.com/uk-news/2026…

Crackdown on tech platforms will go ahead despite US intervention, says No 10

US embassy came out against UK’s proposed under-16 social media ban, which would affect American firms

theguardian.com
95
Coinbase’s new tool can help agents trade and pay for premium research techcrunch.com/2026/06/11/co… via @techcrunch

Coinbase’s new tool can help agents trade and pay for premium research | TechCrunch

Coinbase's agent can use x402 protocol to get access to data and APIs.

techcrunch.com
34
I'm glad Apple isn't hyping up agentic AI (yet) engadget.com/2192124/im-glad…

I'm glad Apple isn't hyping up agentic AI (yet) - Engadget

Apple is more focused on delivering usable features with Siri AI, instead of hyping up agentic AI. That's a good thing.

engadget.com
49
"The manual model breaks": What happens when agents write to production data thenewstack.io/lakefs-agenti… via @thenewstack

"The manual model breaks": What happens when agents write to production data

lakeFS launches an agentic AI data service with isolated sandboxes, branch-scoped credentials, and audit trails to bring governance to autonomous AI workloads.

thenewstack.io
46
Anthropic backtracks on its decision to quietly limit Fable 5's ability to develop LLMs, saying “requests will visibly fall back to Opus 4.8”, after backlash wired.com/story/anthropic-re…

Anthropic Walks Back Policy That Could Have ‘Sabotaged’ AI Researchers Using Claude

The company changed course after researchers spoke out against the policy, which would have covertly limited Claude’s ability to develop competing AI models.

wired.com
76
MassMutual's AI strategy: 12-month contracts, 30% productivity gains, zero lock-in venturebeat.com/orchestratio… via @VentureBeat

MassMutual AI: 12-month contracts, no lock-in

MassMutual caps vendor contracts at 12 months and runs a multi-model architecture — cutting contact center resolution times from 10 minutes to one.

venturebeat.com
62
As the world embraces EVs, the U.S. hits the brakes restofworld.org/2026/us-ev-s…

As the world embraces EVs, the U.S. hits the brakes

Canada and the EU opened their doors to Chinese electric cars this year while the U.S. watched from behind a tariff wall.

restofworld.org
47
Node4 CEO Neil Muller found dead at home after suspected stabbing theregister.com/channel/2026…

Node4 CEO Neil Muller found dead at home after suspected stabbing

MSP says it is 'absolutely devastated' as woman arrested on suspicion of murder

theregister.com
261
Anthropic’s CEO says he doesn’t know if Claude was used in the Iran school strike that killed 120 children thenextweb.com/news/anthropi… via @thenextweb

Anthropic CEO doesn't know if Claude hit Iran school

Dario Amodei said he doesn't know what role Claude played in a strike that killed 120 children at an Iranian school. The AI is embedded in the Pentagon's targeting platform.

thenextweb.com
54
Cheap Iranian drone downed $25 million US Army helicopter—maybe by chance arstechnica.com/gadgets/2026…

Cheap Iranian drone downed $25 million US Army helicopter—maybe by chance

The US military struck Iran again after an Iranian drone’s lucky midair strike.

arstechnica.com
57
Chinese activist in UK told by X that abusive deepfakes do not breach rules theguardian.com/technology/2…

Chinese activist in UK told by X that abusive deepfakes do not breach rules

Apple Peiqing Ni targeted by account portraying her as promiscuous drug addict after posting about Tiananmen Square

theguardian.com
81
Datadog veterans launch AI coding startup Niteshift on a bet against Big AI lock-in techcrunch.com/2026/06/10/da… via @techcrunch

Datadog veterans launch AI coding startup Niteshift on a bet against Big AI lock-in | TechCrunch

AI coding agent startup Niteshift has raised a $7 million seed round from a who's who of angels. It's betting companies will want power over, not lock-in with model makers.

techcrunch.com
54
German court holds Google liable for false AI Overview answers engadget.com/2191469/german-…

German court holds Google liable for false AI Overview answers - Engadget

A recent study found that Google's AI Overviews regularly provide incorrect information and contain facts not supported by cited sources.

engadget.com
46
Databricks wants to kill the "email me a file" problem for AI agent skills thenewstack.io/databricks-op… via @thenewstack

Databricks wants to kill the "email me a file" problem for AI agent skills

Databricks' OpenSharing protocol extends Delta Sharing to cover AI models, agent skills, and unstructured data — now a Linux Foundation project with on-prem storage support.

thenewstack.io
46
Hands-on with Siri AI: successfully executed multistep prompts, understood context well, has strong guardrails, and seems a bit more dispassionate than Gemini theverge.com/tech/947432/sir…

I tried Siri AI, and so far it actually works

Siri might be for real this time.

theverge.com
69
Load more