Capturing Distant Galaxies and Cracking Down on Deepfakes

Emerging Tech Roundup — November 17

The Quantious team’s top picks for timely trending news in the tech world.

This week in tech: NVIDIA Unleashes Next Generation of AI Supercomputer Chips, James Webb Space Telescope Captures the Second-Most Distant Galaxy Ever, Silo AI Unveils New Open Source Language Model for Europe, AI Revolutionizes Weather Forecasting, 3D Generative AI Platform Atlas Emerges to Accelerate Virtual Worldbuilding, YouTube Starts Cracking Down on AI Clones of Musicians, and Amazon’s Astro Robot Gets its First Job as a Security Guard.

NVIDIA’s Next Generation of AI Supercomputer Chips is Here

(Engadget, November 13)

NVIDIA has launched its next-generation AI supercomputer chip, the HGX H200 GPU, based on the “Hopper’’ architecture. This technology offers 141GB of memory at 4.8 terabytes per second, nearly doubling capacity and providing 2.4 times more bandwidth than its predecessor. The chip is designed to enhance deep learning and large language models, such as OpenAI’s GPT-4, and is set to be deployed by major cloud providers for tasks like weather prediction, drug discovery, and quantum computing.

James Webb Space Telescope Captures the Second-Most Distant Galaxy Ever

(Interesting Engineering, November 13)

Scientists utilizing the James Webb Space Telescope have identified the second and fourth-most distant galaxies ever observed — named UNCOVER z-13 and UNCOVER z-12 — situated approximately 33 billion light-years away in the Pandora’s Cluster (Abell 2744). The $10 billion observatory’s spectroscopic data played a crucial role in this discovery, with researchers noting that the galaxies appear larger than others observed at similar distances. Prior to this finding, only three galaxies were confirmed to exist at such extreme distances.

Silo AI Unveils Poro, a New Open Source Language Model for Europe

(Venture Beat, November 13)

Finnish AI startup Silo AI has introduced Poro, an open-source large language model (LLM) geared towards enhancing multilingual AI capabilities for European languages. Poro marks the initial release in a series of planned open-source models intended to cover all 24 official European Union languages. Designed to address challenges in training language models for lower-resourced European languages, Poro utilizes a cross-lingual training approach to leverage data from higher-resourced languages like English.

Courtesy of AI: Weather Forecasts For the Hour, the Week, and the Century

(Tech Crunch, November 14)

Google is leading the charge in the integration of machine learning models for weather forecasting, particularly through DeepMind’s “nowcasting” technology. This approach analyzes precipitation maps to predict weather changes, such as the likelihood of rain, with zero understanding of weather physics. Google aims to dominate in the field of weather forecasting, showcasing the increasing significance of machine learning in delivering accurate and timely predictions.

3D Generative AI Platform Atlas Emerges From Stealth with $6M to Accelerate Virtual Worldbuilding

(Tech Crunch, November 14)

3D generative AI platform Atlas has emerged after two years with $6 million in seed funding, including a $4.5 million round led by 6th Man Ventures and a $1.5 million round led by Collab+Currency. The platform collaborates with game developers and brands to accelerate the creation of virtual worlds, allowing developers to generate detailed 3D models quickly from reference images and text. Developed in-house by a team with expertise in AI, gaming, and design, Atlas aims to streamline workflows and serve as a collaborative design partner.

YouTube is Going to Start Cracking Down on AI Clones of Musicians

(The Verge, November 14)

YouTube has outlined two sets of content guidelines for AI-generated deepfakes, implementing strict rules for the protection of the platform’s music industry partners and a more lenient set for other creators. Creators will be required to label “realistic” AI-generated content, a crucial safeguard for sensitive topics like elections or conflicts. The disclosure will be displayed in video descriptions and on sensitive videos beginning next year, although the platform has yet to define explicitly what it considers “realistic.”

Amazon’s Astro Robot Gets its First Job, Going to Work as a Security Guard

(Geek Wire, November 15)

Amazon has introduced Astro for Business, deploying its robotic Astro as a security guard capable of patrolling and mapping up to 5,000 square feet using HD periscope and night vision. The robot can operate autonomously on pre-programmed routes or manually through the Amazon Astro app, addressing blind spots and areas overlooked by stationary cameras. The business version of Astro starts at $2,349.99, with additional subscription plans available.