I built the solution to this at http://
therundown.ai 🙂
@rowancheung
-
Apple Quietly Acquires AI Startups Ahead WWDC June Announcement
By
–
Yeah, they’ve been pushing tons of research and making AI startup acquisitions without any PR. I’m watching for something big during WWDC in June
-
AI Development Speed Reaches Peak Before Inevitable Slowdown
By
–
& this is the slowest AI development will ever be
-
Maisa Launches Knowledge Processing Unit for Advanced AI Reasoning
By
–
Maisa announced the beta release of its Knowledge Processing Unit.
— Rowan Cheung (@rowancheung) 17 mars 2024
It sets new standards in reasoning, comprehension, and problem-solving by combining the power of LLMs with decoupled reasoning and data processing. https://t.co/6AX0o3pEEsMaisa announced the beta release of its Knowledge Processing Unit. It sets new standards in reasoning, comprehension, and problem-solving by combining the power of LLMs with decoupled reasoning and data processing.
-
Bernie Sanders Proposes Four-Day Work Week Amid AI Automation
By
–
Bernie Sanders proposed a bill establishing a four-day work week in the U.S. with no loss of pay, citing productivity gains from AI and automation.
-
Devin AI Agent Autonomously Solves Coding Issues via Slack
By
–
Cognition’s Devin showed off its autonomous AI agent skills in a new viral X post. The model was able to open a support ticket and exchange Slack messages to solve a coding issue on its own. Wild!
-
India Removes AI Model Approval Requirement After Global Backlash
By
–
India’s Ministry of Electronics and IT revised its AI advisory to remove the requirement for firms to gain government approval before deploying AI models. The move follows major backlash from global entrepreneurs and investors.
-
Google DeepMind’s VLOGGER: AI Avatar Video Generation From Image and Audio
By
–
Google DeepMind researchers just developed VLOGGER.
— Rowan Cheung (@rowancheung) 17 mars 2024
It's a new AI model that can generate talking avatar videos with full upper body motion from just a still image and audio clip.
Lots of use cases here from realism for AI assistants, real-time video dubbing, and more. pic.twitter.com/Q5FsPFHjD5Google DeepMind researchers just developed VLOGGER. It's a new AI model that can generate talking avatar videos with full upper body motion from just a still image and audio clip. Lots of use cases here from realism for AI assistants, real-time video dubbing, and more.
-
Grok-1: xAI Releases 314B Parameter Language Model Weights
By
–
Elon Musk and xAI just released the weights + architecture of Grok-1. It's a massive 314B parameter language model that's 2x the size of GPT-3.5. Big win for collaborative and transparent AI development, and nice to see Elon walking the walk.
-
AI Companions as Gaming Sidekicks: Market Convergence
By
–
I think what's cool here is when agents become sidekicks/companions that play alongside you. We've all seen the numbers for Character AI and companionship-type AI apps. Combining that with gaming just makes sense