AI NEWS YOU MISSED ❗#21 INSEAD 🤖 AI

Robert Maciejko
8 min readApr 14, 2024

--

Suggested by INSEAD-ers. From the 1500+ person global INSEAD 🤖 AI private community, regularly sharing AI-related articles, vids, etc, of interest.

Video Review of Top 5 Stories

Top 5 Stories of the Week

1️⃣ Four New Benchmark AI Models — this week alone

Four cutting-edge AI models were released, from an innovative music generator to advanced LLM and multimodal models setting new benchmarks

Robert Maciejko

2️⃣ But Are AI Models Hitting A Performance Ceiling?

AI influencers Gary Marcus and Ethan Mollick debate whether AI models are hitting a ceiling in performance.

Gary Marcus
Maxime Labonne

3️⃣ AI Chip Wars: New Battles

Recent developments in AI chip technology see major companies like Apple, Intel, Google, and Amazon innovating with new chips to boost AI performance and efficiency, amidst national efforts like the CHIPS Act in the U.S. to enhance semiconductor production and reduce dependency on foreign technology.

Robert Maciejko

4️⃣ AI is still in very early stages

Only about 1% of humans using AI regularly, indicating that AI adoption is still in its nascent stages. OpenAI leads, but it is early.

data tx to Nat Friedman

We have to remember, though, that the Generative AI era has just started. This data is noise.

Exhibit: Robert Maciejko

5️⃣ RSVP This Thursday — April 18 — Can’t Miss AI Interview- Live

INSEAD Prof. Theos Evgeniou interviews MIT legend Tomaso Poggio and Mobileye CEO Amnon Shashua on Half a Century Shaping the AI Frontier — Thursday, 18 April 2024–5.00 pm CET (11 US Eastern)

Listeners may learn about the current and future intersections between AI and neuroscience, and whether traditional computing architectures like GPUs will remain central to AI development.

More upcoming live events

April 24 — VILLE SATOPÄÄMastering Forecasting with INSEAD Professor Ville Satopää — Live Event

INSEADer Content

New INSEAD research paper about business leaders’ attitudes towards Generative AI — INSEAD Prof. Jason Davis

Jason Davis & team

China’s AI Journey: Inside a Global PowerhouseRobert Maciejko

Exhibit: Robert Maciejko

Top Models Update (with DBRX) — by x.com/virattt

Image Generation: Where AI’s hunger for energy comes from

What will humans do if technology solves everything? — Economist

In the article “Welcome to an artificial-intelligence Utopia,” Nick Bostrom explores a world where AI solves all economic tasks and even performs roles traditionally reserved for humans, proposing scenarios ranging from post-scarcity societies to advanced child-rearing AI systems.

  • AI may handle all economically valuable work and potentially parenting, challenging the traditional human roles.
  • This could lead to a “post-scarcity” society where physical resources remain the limit, not economic productivity.
  • As AI evolves, competition and status-driven “positional goods” might still dominate social dynamics, indicating enduring human desires for status and achievement.

Intel unveils latest AI chip as Nvidia competition heats up

Introducing Google Gemini 1.5 Pro

GPT-4 Turbo is now available to paid ChatGPT users

Mistral released Mixtral 8x22B a 141B MoE via magnet link

2023 Europe and US Data, Analytics, and Artificial Intelligence Executive Organization and Compensation Survey

Building a Social Media Algorithm That Actually Promotes Societal Values — Stanford Human-Centered AI

A new camera can undress people almost in real time — to send a message about AI

Why the Chinese Government is sparing AI from harsh regulation — for now

Debunking Devin: “First AI Software Engineer” Upwork Lie Exposed

Open source version of Devin here

“I think seeing agents as “hype” is going to result in people being blindsided as they become more powerful in the coming months. Agents are the explicit goal of OpenAI & Google AI” — Ethan Mollick

How We’ll Reach a 1 Trillion Transistor GPU

New Tech That Asks ‘Are You Sure About Sending a Nude Photo?’

Sam Altman visits UAE this week for talks on investment and global collaboration

Humane AI Pin critique

Claude 3 Opus, Antropic most capable model for highly complex tasks, is now in public preview on Google Cloud’s Vertex AI.

Europe’s A.I. ‘Champion’ Sets Sights on Tech Giants in U.S.

Amazon adds AI heavyweight Andrew Ng to board of directors as Jassy says it may be biggest tech breakthrough ‘perhaps since the internet’

Watsonx Chat, a new interface for interacting with some of the LLMs available in the IBM watsonx.ai studio.

Generative AI Models — Opportunities and Risks for Industry and Authorities — Germany

Adobe’s ‘Ethical’ Firefly AI Was Trained on Midjourney Images

Reasoning still missing in foundation models

Medium clamping down on AI-generated content

Medium email to writers

Rather astounding Assisted Technology coming for Visually Impaired coming

Dotlumen

Orcam: Empowering Accessibility with AI

Other TOP AI Videos of the Week

For a broader selection, check out the INSEAD AI YouTube 🤖 Channel:

What if Dario Amodei Is Right About A.I.?

TL;DW Main points:

  1. AI Development Pace: AI developers experience an accelerated sense of time, often predicting significant advancements and sci-fi-like futures to occur within just a few years, due to exponential growth driven by increased data and computing power.
  2. Societal Impact and Readiness: There’s a notable gap between the rapid advancement of AI technologies and society’s ability to understand and adapt to these changes. This gap could lead to significant social and ethical challenges as AI capabilities increasingly infiltrate everyday life.
  3. Regulation and Ethical Concerns: The conversation highlighted the need for comprehensive regulatory frameworks to manage the development and deployment of AI technologies. This includes addressing ethical concerns like data privacy, misuse of AI in creating bioweapons, and ensuring equitable benefits from AI advancements.

Sam Altman — OpenAI/ChatGPT at Howard University

TL;DW Main points:

  1. Rapid AI Evolution: The pace at which AI is advancing is captivating, with the technology moving at an unprecedented speed.
  2. Ethical Priorities: As an institution committed to social justice and ethics, the discussion focuses on how to ensure that ethical considerations are prioritized in AI development, despite the rapid advancements.
  3. AI’s Potential: AI is recognized as a powerful tool that could lead to significant economic growth and societal improvements, but its successful deployment depends on a collaborative effort involving diverse sectors of society to ensure equitable benefits.

Transforming AI — the Transformer Paper team together in one panel| NVIDIA GTC 2024 Panel Hosted by Jensen Huang

All the authors of “Attention Is All You Need,” the groundbreaking paper that introduced the transformer neural network architecture.

TL;DW Main points:

  1. Introduction of the Transformer Creators: Jensen Huang introduced the inventors of the Transformer model, highlighting their various entrepreneurial ventures post-Transformer. This set the stage for a deep dive into the transformative impact of their work on AI and machine learning.
  2. Discussion on the Transformer’s Impact: The panel discussed the genesis and evolution of the Transformer model, initially designed to enhance machine translation but ultimately revolutionizing multiple AI applications. They highlighted how this model fundamentally changed the approach to processing sequential data, leading to advancements across various fields including natural language processing and computer vision.
  3. Future of AI and Transformer Technology: The inventors shared insights on potential future enhancements for the Transformer model, including more efficient computation and adaptive algorithms that could handle more complex reasoning tasks with less data. They explored how ongoing innovations could lead to more sophisticated models that integrate learning from user interactions, further blurring the lines between human and machine cognition.

Intel Vision 2024 Keynote (Replay) — Intel CEO Pat Gelsinger and special guests

TL;DW Main points:

  1. Intel’s Gaudi 3 Accelerator: Intel launched the Gaudi 3 AI accelerator, boasting four times the AI computing power of its predecessor, better memory bandwidth, and enhanced networking capabilities. It’s designed to significantly boost AI training performance across various applications.
  2. AI Everywhere Vision: Intel is pushing the vision that every company will become an AI company, emphasizing the need for scalable and flexible infrastructure that supports privacy, security, and fast data processing. Their strategy includes delivering technology that adapts to different business needs and accelerates digital transformation.
  3. Future-Proof Innovations: Alongside Gaudi 3, Intel is working on next-generation AI solutions like Lunar Lake and advanced Ethernet technologies tailored for AI workloads. These developments are aimed at providing high-performance, cost-effective options for handling large-scale AI operations and enhancing data connectivity and security in AI deployments.

Past editions — AI NEWS YOU MISSED

AI News You Missed

49 stories

SUBSCRIBE TO GET UPDATES IN YOUR MAILBOX

Follow Robert Maciejko on LinkedIn or X (Twitter)

--

--

Robert Maciejko
Robert Maciejko

Written by Robert Maciejko

Entrepreneurial Leader & International Change Driver who delivers. Co-founder of the 1500+ strong global INSEAD AI community. Opinions are personal.

No responses yet