←
Scale2,323 posts
Scale avatar

Scale

@scale_AI

making AI work

◷ Joined July 2016▧ 598 Photos and videos
Francis deSouza avatar
⟳ Scale retweeted
Francis deSouza@fdesouza·Nov 16

Hospitals need proof an AI agent can do the work before it touches a patient record. Today at @HLTHEVENT we're launching 5 agents for hospital operations: → Scheduling → Order guidance → Medical record review → Safety & incident investigation → Clinical documentation & quality review This builds on our work with @MayoClinic, where our agents turn the tens of millions of pages of outside records Mayo gets each year into something a physician can actually use. One Mayo physician says it saves 5–30 minutes of prep per visit. scale.com/blog/scaleforhospi…

Scale avatar
Scale@scale_AI·Nov 16

We’re building the future of reliable AI in healthcare, starting with the operations that keep hospitals running. Introducing Scale for Hospital Operations: five agents built to strengthen the care teams behind every patient.

0:51
Cesar Yazbeck avatar
⟳ Scale retweeted
Cesar Yazbeck@CesarYazbeck·Nov 13

I’m joining @scale_AI as Head of Healthcare & Life Sciences. As a medical doctor who has spent the past decade working in the industry, I believe we’re at an inflection point and it’s time to deliver real impact for healthcare teams and patients. Why I joined Scale and what we’re building for healthcare: scale.com/blog/why-i-joined-…

Francis deSouza avatar
⟳ Scale retweeted
Francis deSouza@fdesouza·Nov 12

Qatar’s AI Smart Legislative Advisor cut legislative review from ~35 days to ~5. @scale_AI just signed two more agreements with Qatar’s Ministry of Justice and @MCITQatar: Sakk for notarization, Shoor for state-case legal research. gulf-times.com/article/73790…

Scale avatar
Scale@scale_AI·Nov 11

Service doesn't end when the uniform comes off. Our veteran colleagues bring invaluable experience and leadership to Scale, helping us build trusted AI for our nation's defense. This Veterans Day, we honor our veterans and all who have served. Thank you for your service.

Post attachment
Francis deSouza avatar
⟳ Scale retweeted
Francis deSouza@fdesouza·Nov 11

Many veterans never stop serving. Some of our colleagues at @scale_AI served in uniform. They know what it means to depend on a system when failure isn’t theoretical. That shapes the AI we build for national defense. To our veterans, and to all who have served, and their families: thank you.

Post attachment
Francis deSouza avatar
⟳ Scale retweeted
Francis deSouza@fdesouza·Nov 10

Congrats to @DarioAmodei and the @AnthropicAI team on Fable 5.2 results! We had the opportunity to evaluate the model ahead of launch for our Remote Labor Index benchmark, and the results are remarkable. Just months ago, no model had crossed 10% on RLI. Today, Fable 5.2 scores over 40%, nearly doubling Opus 5.5. A major leap forward in AI's ability to tackle real-world work.

🥇 @AnthropicAI's Fable 5.2 takes the top spot on our Remote Labor Index with @CAIS, our leaderboard which compares AI agents' work head-to-head with work by human professionals on real-world remote tasks. The model successfully automates 41.3% of projects, nearly doubling the performance of the next-best model Opus 5.5, a major leap in AI's ability to handle real-world work. A closer look at where the biggest jumps are coming from 🧵

0:06
Francis deSouza avatar
⟳ Scale retweeted
Francis deSouza@fdesouza·Nov 9

Sovereign AI isn’t one decision. It’s a set of choices a country keeps making: where its data lives, which models it relies on, which cloud it trusts in production. This week at MWC Doha, @scale_AI and @Oracle published a joint reference architecture for running our GenAI Portfolio on Oracle’s in-country cloud. It builds on our multi-cloud blueprints for Qatar. Another path to keeping data and models under their own control. scale.com/partnership/oracle

Francis deSouza avatar
⟳ Scale retweeted
Francis deSouza@fdesouza·Nov 5

Today we're publishing a major RL environments update: 3,300+ tasks across tool use and computer use, already running with customers. Some tasks in our newest RL environments take a human team 3 weeks and 500+ tool calls. scale.com/rlenvironments

Francis deSouza avatar
⟳ Scale retweeted
Francis deSouza@fdesouza·Nov 5

Congratulations to Qatar’s Council of Ministers Secretariat General and @MCITQatar on winning Best Use of AI at the 7th GCC Digital Government Award. Every new law has to fit with every existing one. The Smart Legislative Advisor surfaces the conflicts; legislators decide. Proud @scale_AI helped build it. qna.org.qa/en/news/news-deta…

Post attachment
Francis deSouza avatar
⟳ Scale retweeted
Francis deSouza@fdesouza·Oct 29

@scale_ai is teaming with @huggingface to make robots better at long, multi-step tasks, and we’re open-sourcing all of it! Today @scale_AI is launching FineART-VLA, a model that helps robots complete longer tasks, building on our first open-source robotics dataset. Built with @huggingface 's LeRobot team, this is Scale’s first open-source robotics release. Thanks to @Thom_Wolf + team for the partnership. The momentum on open-source robotics is picking up steam, and high-quality data will determine how far it goes. Go build! labs.scale.com/blog/fineart

Scale avatar
Scale@scale_AI·Oct 29

Better robots start with better data. FineART pairs 500K+ hand-annotated robot steps with an open model built to learn from them, helping close a critical gap in teaching robots to complete complex, multi-step tasks.

In collaboration with @HuggingFace, we’ve released the largest open robot dataset, alongside their first open System 1-System 2 model that can reason over step-by-step actions. FineART is built with 40K+ demonstrations across 151 tasks and 500K subtask labels. Together, the dataset and model were designed for each other, and both are available on @LeRobotHF where robotics developers work today.

2:43
Francis deSouza avatar
⟳ Scale retweeted
Francis deSouza@fdesouza·Oct 28

DualChem, a new benchmark from @scale_AI, tested 29 frontier models on dual-use chemistry questions: ones that could serve either legitimate scientific research or harmful ends. They frequently flagged the risk, then answered anyway. Recognition is not refusal. DualChem is part of our research collaboration with @NIST’s Center for Advancing Innovation and Standards for Super Intelligence (CAISSI). Governments need independent testing capacity they don’t have today. @madialder covered the findings at FedScoop: fedscoop.com/scale-chemistry…

Francis deSouza avatar
⟳ Scale retweeted
Francis deSouza@fdesouza·Oct 28

NEWS: @scale_AI is putting down roots in Qatar. Today we signed an agreement with @UDCQatar to establish a long-term home in Doha. Our team in Doha has grown to nearly 100 people. The office at The Pearl builds on our five-year partnership with @MCITQatar, where our team is making healthcare, justice, education and tourism services more reliable for the people who depend on them. We’ve also trained nearly 6,000 government employees. Qatar’s AI ambitions depend as much on its people as on its technology, and building that literacy and confidence matters as much as anything we deploy. Thank you to MCIT for its leadership, and to UDC for helping us build a lasting home. scale.com/blog/scales-long-t…