Meta's Adam Mosseri Warns AI Token Spend Could Be Capped Per Engineer
Meta's Adam Mosseri said AI token budgets could be capped per engineer within a year or two, citing rising costs that nearly pushed the company to billions in 2026.
Archive
Browse all published articles.
ARCHIVE
-
-
DeepMind CEO Pushes for Independent AI Standards Body
Google DeepMind CEO Demis Hassabis proposed a new regulatory body to oversee frontier AI models, modeled after FINRA, in a July 14 post.
-
Google Faces AI Training Lawsuit From Major Publishers
A group of publishers and authors has sued Google, alleging the company trained its Gemini AI on copyrighted works without permission.
-
AMD Introduces LogsLop for Efficient Log File Summarization
AMD released LogsLop, a tool that reduces repetitive log files to 34 representative messages, helping users quickly identify critical errors.
-
ScienceSoft Builds HIPAA-Compliant AI Voice Scheduler on AWS
ScienceSoft's AI voice scheduler, built on AWS, helps healthcare providers streamline scheduling while ensuring HIPAA compliance. The market is valued at $260 million in 2023 and is projected to exceed $1.2 billion by 2030.
-
Flo Health Uses Amazon Bedrock to Cut Medical Review Time by 60%
Flo Health reduced medical content review time by 60% using Amazon Bedrock, tripling content throughput without expanding its medical team.
-
Amazon Introduces Nova Act for Automated UX Testing
Amazon unveiled Nova Act, a multimodal model that analyzes web interfaces through vision and action, enabling scalable UX testing for digital platforms.
-
Amazon Introduces QA Studio for Agentic QA Automation with Nova Act
Amazon's QA Studio streamlines regression testing with batch execution and CI/CD integration, supporting up to 20 tests running concurrently.
-
Thrad.ai Uses Strands Agents and Amazon Bedrock for AI Advertising
Thrad.ai leverages multi-agent systems to automate social intelligence for AI advertising, scoring prospects with 100-point scale based on signal triangulation.
-
NVIDIA Introduces Nemotron Open Models for Customizable AI
NVIDIA announced open models like Nemotron, enabling enterprises to customize AI for specific needs, with examples showing cost savings and improved accuracy.
-
LLM Security Flaws Exposed in AI Jailbreaking Experiment
Researcher Dave Kuszmar bypassed safety measures in major LLMs, including Google Gemini and OpenAI's GPT-4o, to obtain detailed instructions on dangerous activities.
-
Superhuman Launches Improved Auto-Draft Feature for Email
Superhuman's new auto-draft feature, tested by the company's co-founder, generates replies that require minimal editing, with 40% of drafts sent within a day.