ODSC’s AI Weekly Recap: Week of February 23rd

Open Data Science Blog Recap

A new open-source Python package looks to push for accuracy and reliability in the outputs of large language models. (Source)

LAION has introduced, BUD-E, an open-source voice assistant that is able to run on a gaming laptop and does not require an active internet connection. (Source)

OpenAI’s latest model, Sora a text-to-video model that aims to reshape the creation and interaction of visual content by leveraging the capabilities of AI. (Source)

The University of Pennsylvania School of Engineering and Applied Science unveiled its Bachelor of Science in Engineering in Artificial Intelligence. (Source)

In an unprecedented surge, Nvidia’s market capitalization galloped past the $2 trillion mark post-Thursday’s trading session (Source)

The Biden administration has stepped into the ongoing debate surrounding the openness of artificial intelligence systems. (Source)

AI News Highlights

School leaders in Cumberland County are using Artificial Intelligence to help students deal with mental health issues with a chatbot named Kiwi the Llama.

Google is setting up a new hub in France dedicated to the artificial intelligence (AI) sector, which will be inaugurated on Thursday.

A new initiative by metaLAB at Harvard teaches educators how to integrate AI into their pedagogy.

Shareholders of Apple are currently eagerly awaiting new AI news related to products and services.

A growing number of Michigan small businesses are finding success using artificial intelligence in their daily operations, according to a new survey.

Lead researcher on the MinePlanner project Steven James observed that future AI models will need to tackle messy problems and he hopes to test how well they deal with unfamiliar scenarios by getting them to play Minecraft.

The NBA has unveiled NB-AI, technology that the league says will “personalize the live game viewing experience,” at the annual All-Star tech summit Friday.

Google has introduced Gemini 1.5, a model that delivers dramatically enhanced performance, with a breakthrough in long-context understanding across modalities.

Speaker of the House of Representatives Mike Johnson and Democratic leader Hakeem Jeffries launched a bipartisan Task Force on Artificial Intelligence.

Two tech CEOs scrambling to produce more of the sophisticated chips needed for artificial intelligence.

New research suggests artificial intelligence agents can develop trust similar to that of humans.

China’s Rush to Dominate A.I. Comes With a Twist: It Depends on U.S. Technology

Trending AI Open Source Projects

• Face Recognition Recognize and manipulate faces from Python or from the command line with the world’s simplest face recognition library.
• google-indexing-script Use this script to get your entire site indexed on Google in less than 48 hours. No tricks, no hacks, just a simple script and a Google API.
• Lepton: a lean code snippet manager powered by GitHub Gist.
• Opengfw is a flexible, easy-to-use, open-source implementation of GFW on Linux that’s in many ways more powerful than the real thing. It’s cyber sovereignty you can have on a home router.
• Fabric is a high-level Python (2.7, 3.4+) library designed to execute shell commands remotely over SSH, yielding useful Python objects in return. It builds on top of Invoke (subprocess command execution and command-line features) and Paramiko (SSH protocol implementation), extending their APIs to complement one another and provide additional functionality.

Research From Around The Globe

• Non-Exchangeable Conformal Language Generation with Nearest Neighbors — Paper
• Gemma — Google’s new open LLM — Hugging Face
• BootsTAP: Bootstrapped Training for Tracking-Any-Point- Paper
• Large Language Models as an Indirect Reasoner: Contrapositive and Contradiction for Automated Reasoning — Paper
• Lumos: Empowering Multimodal LLMs with Scene Text Recognition — Paper
• Graph Mamba: Towards Learning on Graphs with State Space Models — Paper
• A phase transition between positional and semantic learning in a solvable model of dot-product attention — Paper
• Understanding the Reasoning Ability of Language Models From the Perspective of Reasoning Paths Aggregation — Paper
• CodeIt: Self-Improving Language Models with Prioritized Hindsight Replay- Paper
• Direct Language Model Alignment from Online AI Feedback — Paper
• Feedback Loops With Language Models Drive In-Context Reward Hacking — Paper
• Non-Exchangeable Conformal Language Generation with Nearest Neighbors — Paper
• An Interactive Agent Foundation Model — Paper

Start-Up Funding News

San Jose-based Recogni, a provider of vision-based perception processing for autonomous driving platforms with high computing, low latency, and low power usage., raised north of $200M in a Series C round.

Paris-based Bioptimus, a developer of a universal AI foundation model for biology., raised north of $35M in a Seed round.

California-based Athos Therapeutics, developer of AI-based Precision Therapeutics for patients with Inflammatory Bowel, other Autoimmune Diseases, and cancer raised $54M in a Series B round.

New York-based Qloo, an AI-based cultural platform to predict consumer taste data and recommendations for companies in various sectors, raised $57M in the Series C round

Israel-based Bria, tackling the challenges of implementing Visual Generative AI for commercial use across all departments — business, product, and technology, raised $26.5M in Series A.

New AI Skills Await!

Take flight with your team, and save 40%

Upskill Your Team!

Blogs or News to Share?

If you have any blogs, industry news, product launches, job postings, fundraising news, or open-source launches to add to our newsletter then please send them to: blogs@odsc.com.