AI NEWS YOU MISSED ❗#22 INSEAD 🤖 AI
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 Meta’s battle with ChatGPT and other AI rivals begins now
Meta’s AI assistant is being put everywhere across Instagram, WhatsApp, and Facebook. Meanwhile, the company’s next major AI model, Llama 3, has arrived.
🥁 Llama3 is out 🥁 — Yann LeCun
Welp folks, we have gpt-4 at home
2️⃣ International AI: Microsoft to invest $1.5bn in Abu Dhabi AI group G42
Reflects growing tech collaboration between the UAE and the US at the expense of China
3️⃣ AI Index: Comprehensive 500 page “State of AI” from Stanford HAI is out.
See the State of AI in 13 Charts
Key conclusions are:
- AI beats humans on some tasks but not on all.
- Industry continues to dominate frontier AI research.
- Frontier models get way more expensive.
- The US leads China, the EU, & the U.K. as a leading source of top AI models.
- Robust and standardized evaluations for LLM responsibility seriously lacking.
- Generative AI investment skyrockets.
- AI makes workers more productive & leads to higher-quality work.
- Scientific progress accelerates even further, thanks to AI.
- The number of AI regulations in the United States sharply increases.
- People across the globe are more aware of AI’s potential impact — & more nervous.
Major tech stocks like NVIDIA have fallen sharply of late. Quarterly reports of AI players, including Meta, Alphabet, and Microsoft, coming this week will give more guidance.
5️⃣ AI Music Generation: Viral Drake track that features AI voices of Tupac and Snoop Dogg.
Unclear if Tupac’s crew or Snoop Dogg knew.
Upcoming live events — INSEADer lead
April 24 — INSEAD Prof. VILLE SATOPÄÄ — Mastering Forecasting with INSEAD Professor Ville Satopää — Live Event
May 15 — INSEAD Prof. Felipe Monteiro— Beyond the Digital Core: GenAI’s Impact on Traditional Industries
May 29 — INSEAD Prof. Theos Evgeniou — Open Source AI, Ecosystems & National Adaptation: A View from the UAE
INSEADer Content shared in INSEAD AI groups
Video replay of Interview by Prof. Theos Evegeniou of MIT’s Tomaso Poggio and Mobileye CEO Amnon Shashua is coming
Global Business Leaders on Generative AI with INSEAD Professor Jason Davis
INSEAD Professor Jason Davis surveyed INSEAD Alums on Generative AI. In this interview with INSEAD 🤖 AI co-founder Robert Maciejko, he shares the results.
Topics covered include:
➡️ Generative AI (Gen AI) use by surveyed business leaders.
➡️ Sentiment towards GenAI.
➡️ Predominant concerns.
➡️ Who stands to gain more from GenAI — organizations or individuals
➡️ Differences in perceptions based on age and geographic location
➡️ Predictions for AGI
AI In Business: What Do Business Leaders Need To Know About AI? — With INSEAD Prof. Villa Satopää
A corridor poster from the Massachusetts Institute of Technology Center for Brains, Minds and Machines — Probably tons of future PhD/research projects in here —INSEAD Prof. Theos Evgeniou
Anyone else has finished reading “Burn Book: A Tech Love Story” by Kara Swisher? It’s 🔥🔥🔥😎 — INSEADer Andreia Domingues
Artificial Integrity — INSEADer Hamilton Mann
AI Agents Are Starting to Change the Game at Work — INSEADer Robert Maciejko
Further suggested reading from the INSEAD AI community this week
Welcome to the era of AI nationalism — Economist
The Smartest Way to Use AI at Work — WSJ
How People Are Really Using GenAI — HBR
People are finding all sorts of ways to use generative AI in their daily lives and at work
101 real-world gen AI use cases featured at Google Cloud Next ’24
Meet the woman who transformed Sam Altman into the Avatar of AI
Launch of Mentee Robotics to the public and unveiling the MenteeBot — Amnon Shashua
Return to reality — get Google results pre-2023 to avoid AI sources — Elon Musk
An uncensored model with no guardrails
Multi-agent collaboration — Andrew Ng
Transcript: Ezra Klein Interviews Dario Amodei
Can Large Language Models “Abstract”? — Amnon Shashua
Leave No Context Behind: Efficient Infinite Context Transformers with Infini-attention
A nice and simple guide to getting started with the Gemini 1.5 Pro in Google AI Studio — Google’s Logan Kilpatrick
Tested a good amount of LLama 3 + other models for production use cases with agents: — State of the Art
Gen AI As A Game-Changer in Bank Risk Management
AI editing tools are coming to all Google Photos users
Amazon is filled with garbage ebooks. Here’s how they get made
Quantum Computing: Eric Schmidt Insights
16-Week Online Skills Bootcamps to position you for your first tech role
Machine Learning and Artificial Intelligence courses at MIT
Small (open-source) VLMs — Vision-text-models running as locally
Wayve — Pioneering a new way to solve self-driving with Embodied AI
A patient’s digital twin is a comprehensive forecast of their future health
Silicon Valley leader posted 12 bold predictions — Vinod Khosla
Generative AI: Opportunities and Impact — May 3 Singapore
Your Brain Waves Are Up for Sale. A New Law Wants to Change That
The Generative AI Era Has Just Started
Google restructures finance team as part of AI shift
GPT-4 Can Exploit Most Vulns Just by Reading Threat Advisories
Forget RAG and feed it into a huge context window model like Gemini 1.5 Pro
Pressure Testing GPT-4–128K With Long Context Recall
Time for the open-source AI robots revolution — Thomas Wolf, Hugging Face
Fun:
Other TOP AI Videos of the Week
For a broader selection, check out the INSEAD AI YouTube 🤖 Channel:
EP59: Unhinged Meta Llama 3 *Special Edition*
TL;DW summary:
- Introduction and Capabilities of Llama 3: The episode begins with an exploration of Meta’s newly released Llama 3, an open-source large language model that boasts improved performance and creativity. The hosts highlight a song generated by Llama 3 to showcase its creative applications.
- Technical Specifications and Developer Appeal: Llama 3 is discussed in detail, including its variants with different parameter sizes (7B, 40B, and an upcoming 400B). The model’s open-source nature and its integration into Meta’s popular platforms like Instagram and WhatsApp are particularly noted as potentially appealing to developers looking to easily implement and scale AI-driven features.
- Performance Benchmarks and Competitive Edge: The discussion includes a comparison of Llama 3’s performance against other major models such as OpenAI’s GPT-4. The hosts note Llama 3’s competitive benchmarks, emphasizing its lower false refusal rates and its potential to outperform existing models in creativity and user interaction.
- Integration, Licensing, and Developer Considerations: The episode covers how Llama 3 can be integrated into various applications, with a mandatory licensing requirement that derivatives must credit Llama 3. This aspect is debated in terms of its impact on developer freedom and the broader AI development community.
- Market Dynamics and Future Prospects: There’s significant discussion on the strategic positioning of Llama 3 in the AI marketplace, particularly in relation to OpenAI’s offerings. The hosts speculate on the future enhancements of Llama 3 and discuss how its open-source model could shift competitive dynamics, offering developers a robust alternative to OpenAI’s more closed ecosystem.
All New Atlas | Boston Dynamics
“The next generation of humanoid robots — a fully electric Atlas robot designed for real-world applications. The new Atlas builds on decades of research and furthers our commitment to delivering the most capable, useful mobile robots solving the toughest challenges in industry today: with Spot, with Stretch, and now with Atlas.”
The Rise of The Machines: John Etchemendy and Fei-Fei Li on Our AI Future | Uncommon Knowledge
TL;DW summary:
- Definition and Scope of AI: Fei-Fei Li and John Etchemendy discuss the broad and evolving definition of artificial intelligence, which encompasses data analysis, pattern recognition, decision-making, and learning across various domains such as language, vision, and robotics.
- Historical Milestones in AI: The discussion highlights significant milestones in AI, like the victories of IBM’s Deep Blue and Google’s AlphaGo over human champions in chess and Go, respectively. These examples illustrate AI’s progression from following programmed instructions to learning and making decisions based on data patterns that are not immediately obvious to humans.
- Ethical and Social Implications: The conversation addresses the potential dangers and ethical dilemmas posed by AI, comparing its transformative impact to historical technological advances like nuclear technology. The dialogue emphasizes the need for careful management and ethical considerations in AI deployment.
- AI’s Impact on Employment: The speakers explore AI’s potential to automate tasks within jobs, reshaping the labor market and job descriptions rather than simply eliminating employment. They advocate for proactive adaptation to these changes through education and policy adjustments.
- Regulatory and Policy Considerations: Etchemendy and Li stress the importance of appropriate regulation and policy to guide AI development and integration into society. They argue for a balanced approach that fosters innovation while addressing potential risks, highlighting the role of academia and government in cultivating a beneficial AI ecosystem.
First joint interview of OpenAI CEO Sam Altman and COO Brad Lightcap
TL;DW summary:
- Conservative vs. Progressive AI Strategies: Many startups surprisingly opt for conservative AI strategies that don’t anticipate significant improvements, potentially risking obsolescence as AI technology evolves rapidly.
- Foundational Insight of OpenAI: OpenAI was founded on the unique insight that deep learning’s effectiveness significantly scales with size, a concept that was not widely accepted at the time, indicating a major gap in the AI research landscape.
- Disruption of AI Advancements: OpenAI’s continuous improvements could unexpectedly disrupt startups that do not account for rapid advancements in AI, highlighting the importance of forward-thinking in technology development.
- Research-Driven Business Model: OpenAI’s focus on integrating research with business is critical for fostering innovation but presents challenges as the company scales, emphasizing the need for a balance between innovation and commercial pressures.
- Iterative Deployment for Ethical AI: OpenAI promotes the iterative deployment of AI technologies to facilitate societal adaptation and ethical considerations, a strategy that is surprisingly rare yet essential for the responsible development of impactful technologies.
Past editions — AI NEWS YOU MISSED
SUBSCRIBE TO GET UPDATES IN YOUR MAILBOX
Follow Robert Maciejko on LinkedIn or X (Twitter)
