Category: Artificial Intelligence

Artificial Intelligence, Yogi Nelson

WHAT IS THE ARTIFICIAL INTELLIGENCE ALLIANCE?

Yogi Nelson

Namaste Yogis. Welcome to the Blockchain & AI Forum, where your technology questions are answered, mostly correct! Here no question is too mundane. As a bonus, a proverb is also included. Today’s question comes from Butch, in Arkansas, USA. Butch wants to know, what is the Artificial Intelligence Alliance?

Butch, you came to the right place. The Artificial Intelligence Alliance (AI Alliance) is an organization with an impressive membership roster. The AI Alliance consists of technology creators, developers and adopters collaborating to advance safe, responsible AI rooted in open innovation, according to their official statement. Who are the members? Click the link here: https://thealliance.ai/

AI Alliance membership is diverse and as expected concentrated in the tech sector. There are numerous prestigious universities, including U.C. Berkeley, Dartmouth, and the University of Texas. Moreover, the AI Alliance includes foreign universities in Japan and one in Dubai, exclusively for AI. Holy, big brain, Batman! Not surprisingly AI Alliance members are generally from the business community. For example, you will find Meta, Intel, IBM, Oracle, and many other household corporate names.

What does the AI Alliance do? The AI Alliance has three focus areas. Let’s take one at a time.

Responsible and Scalable AI. AI Alliance has set a goal of creating benchmarks, tools, and methodologies to ensure and evaluate high-quality and safe AI. The AI Alliance will also be an advocate for AI safety and security.

Open Foundation Models. The AI Alliance says it will focus on responsibly advancing the ecosystem of open foundation models with diverse modalities, including highly capable multilingual, multi-modal, and science models that can help address society-wide challenges in climate, education, and beyond.

Diverse AI Hardware. AI Alliance claims its members will collaborate on the benchmarking, optimization, and adaptation of AI workloads to advance a diverse set of hardware. AI Alliance reports it will focus on scalability, platform adaptation, and advanced power, energy, and carbon modeling. Benchmarks and metrics developed for model inference, fine-tuning and energy consumption of large-scale AI workloads will also be contributed to the open-source community.

Butch, you ask in what way do the members contribute? I’ll give you three of the eight ways AI Alliance members can contribute. We start with a pledge to deploy benchmarks, tools, and other resources that enable the responsible development and use of AI systems at global scale, including the creation of a catalog of vetted safety, security, and trust tools. Second AI Alliance members agree to responsibly advance the ecosystem of open foundation models with diverse modalities, including highly capable multilingual, multi-modal, and science models that can help address society-wide challenges in climate, human health, and beyond. And third, AI Alliance members say they will support global AI skills building, education, and exploratory research. Nothing short of ambitious!

The actual working of the AI Alliance is fascinating and deserves a short explanation. The AI Alliance will divide the work and its members into groups with specific tasks. To direct the work, the AI Alliance announced it will establish a governing board and technical oversight committee. The board and technical committee will establish overall project standards and guidelines. There is no mention of an Executive Director, dedicated staff, nor explicit political activities, yet. However, I am confident the AI Alliance will eventually establish a Washington, D.C. presence to lobby the executive branch, Congress, and regulators. That’s how it works, folks!

Butch, I hope I answered your question. Remember what they say in Italy: When you finish the game, the king and pawn end up in the same box.

Until next time,

Yogi Nelson

Artificial Intelligence

WHAT IS THE PAUL ALLEN INSTITUTE FOR ARTIFICIAL INTELLIGENCE

Yogi Nelson

Namaste Yogis. Welcome to the Blockchain & AI Forum, where your technology questions are answered, mostly correct! Here no question is too mundane. As a bonus, a proverb is also included. Today’s question comes from in Mark, Naperville, USA. Mark wants to know, what is the Paul Allen Institute for Artificial Intelligence?

Mark, you came to the right place. Let’s take a moment first to understand, who was Paul Allen. Mr. Allen made his fortune as a Microsoft co-founder. Perhaps his name sounds familiar from the Microsoft connection? Allen passed away in 2018 at age 65. Too young! Lol! Before Allen’s passing, he established three research institutions, including the AI research.

Paul Allen AI Institute, commonly known as AI2, was established in 2014. AI2 is in Seattle and Tel Aviv, Israel. The mission of AI2 is “…to contribute to humanity through high impact AI research and engineering.” Let’s dive into what AI2 produces–it generates plenty—beginning with Aristo.

Aristo. Aristo was AI2’s first project. The concept powering Aristo was to create a “Digital Aristotle” capable of answering users’ questions, and ultimately holding a discussion. Aristo was designed to understand elementary science by reading relevant text. In 2019 Aristo scored over 90% on a collection of multiple-choice, eight grade science exams. Better than me, lol! AI2 readily admits Aristo remains incapable of understanding science in a meaning way. Holy ancient Greek, Batman!

Semantic Scholars. Semantic Scholar accelerates scientific breakthroughs by helping scholars locate and understand key research results, make important connections, and overcome information overload. AI2 says Semantic Scholar is an AI-based search engine for scientific literature, including more than 200 million papers covering all scientific disciplines. It draws 10M users monthly. Semantic Scholar can decode PDF files to extract and understand the full-text content beyond titles, authors, and references to include figures and concepts. Semantic Scholar played a leading role in creating the Covid-19 open research dataset.

AllenNLP. AllenNLP executes groundbreaking research in natural language processing, including improving systems’ performance, accountability, and advancing scientific methodologies for evaluating and understanding systems.

PRIOR. PRIOR is the computer vision research team. The goal is advanced computer vision leading to AI systems that can perceive the world, reason about it, and act within it. PRIOR has three research priorities: 1) embodied AI systems capable of interacting with objects, humans, and other agents in the real world; 2) vision and language multi-modal systems that can perform tasks that require an understanding of visual and language data; and 3) computer vision for the common good, e.g., a platform that can detect illegal fishing vessels via satellite imagery.

Green AI. Green AI is dedicated to reducing the cost of training and using AI systems—less expensive AI systems are accessible to more practitioners and contribute less to climate change.

Mosaic. Mosaic was formed in 2018 with the goal of building AI with common sense. Common sense is the ability to acquire and use every day experiential knowledge. Obviously, Ais lack commonsense. Good luck with this one. Lol.

AI Environment. Applied AI for climate modeling and wildlife protection.
• Earth Ranger: Software that aids ecologists and wildlife biologists make more informed wildlife conservation decisions.
• Skylight: Helps reduce illegal, unreported, and unregulated fishing through technology that provides actionable intelligence for maritime enforcement.
• Climate Modeling: The goal is to improve the world’s understanding of climate change, its effects, and what can be done now.
• Wildlands: Machine learning to support wildland fire management.

We conclude with a Norwegian proverb: Behind the clouds, the sky is always blue.

Until next time,

Yogi Nelson