Hey Everyone,

Welcome to the Generative AI Brief, TGAB Issue #20 already. This newsletter is for hardcore AI enthusiasts who like to read. If you don't like to click on links, you might not enjoy this.

But, I hope you appreciate links and reading about A.I. as much as I do! 🚀

This post is a quick link tool to completely get up to date (provided you have time to read) on what occured in A.I. during the last week to ten days. Week of October 30th, 2023.

​

• To keep getting these Newsletters star me and take me out of Google Promotions tab and “Move to Primary”. Bookmark this post to come back to read on the Weekend.

​

If this is of value to you, support the channel. Get exclusive access to deep dives of A.I. Supremacy, my flagship Newsletter.

​

​Subscribe now​

​

For the best reading experience, read this on a web browser.

​

This issue is made free by Aircall, I cannot obtain a minimum living wage by subscriptions in full yet, which means I have to embrace sponsors. By clicking on and signing up for Sponsors you are supporting the channel.

---

💓My Sponsor Company🚀

Introducing Aircall AI - Simple Yet Powerful AI Features for Voice

Why use Aircall’s conversation AI?

• Less admin, more time - save 50% of time spent on call reviews

​Register Today​

​

---

Advertisers & AI Startups: Choose one of my Newsletters to Sponsor if you think we might be a good fit.

​

💬Introduction for Today🎯

• The AI summit in the UK was supposed to be a big political and technological moment for the future of x-risk. Instead it felt like a PR moment for politicians and warnings and feuds by various A.I. personalities.
• The A.I. News cycle was not very dynamic this week.
• Decent week for A.I. appears again though.

​

---

​A Tweet on X (Elon Musk)

​

📢Top News 📺

​

UK invests $273 million in AI supercomputer as it seeks to compete with U.S., China​

​

Microsoft: Copilot could drive Billions in revenue on the back of GPT-4 and GPT-5 (CNBC, projection report)

​

Together.AI:RedPajama-Data-v2: an Open Dataset with 30 Trillion Tokens for Training Large Language Models​

​

Andrew Ng: Google Brain founder says big tech is lying about AI extinction danger​

​

Yann LeCun: AI one-percenters seizing power forever is the real doomsday scenario, warns AI godfather (Insider)

​

OpenAI: Frontier risk and preparedness (Official)

​

​AI Doomers Take Center Stage at the UK’s AI Summit (Bloomberg)

​

​Day 1 of UK's AI safety summit. Here's what you need to know (Ben’s Bites)

​

A report by Ernst & Young (EY), one of the Big Four accounting firms, on the global AI regulatory landscape.

​

OpenAI’s ChatGPT app revenue shows no signs of slowing, but some other AI apps top it ​

​

🔗Partnership & Announcements🏭

​

​

​G7 to agree AI code of conduct for companies (Reuters)

​

​Rishi Sunak confirms AI Safety Institute (TechCrunch)

​

​Google rolls out generative AI tools for product images in the U.S.

​

Amazon:New AWS service lets customers rent Nvidia GPUs for quick AI projects​

​

​

📰Noted Newsletters📬

​

​Microsoft Copilot 365 Cannot Really Fail, Here's Why (AI Supremacy)

​

Andrew Ng: Sensationalist claims that AI could bring about human extinction will cause serious harm.​

​

​Attenuating Innovation (AI) (Stratechery) ​

​

---

📝 Papers 🗏

​

​GLM-130B: An Open Bilingual Pre-trained Model (Oct 5th) - Context (China)

​

​Text Embeddings Reveal (Almost) As Much As Text (Oct 10th)

​

​DreamCraft3D: Hierarchical 3D Generation with Bootstrapped Diffusion Prior (October 25th)

​

Apple: Matryoshka Diffusion Models (MDM) (Oct 23rd, Apple) - Additional Context​

​

Tencent: Progressive Conditional Diffusion Models (PCDMs) - Advancing Pose-Guided Image Synthesis with Progressive Conditional Diffusion Models (Oct 10th)

​

​

​Zephyr: Direct Distillation of LM Alignment​

​

VideoCrafter1: Open Diffusion Models for High-Quality Video Generation​

​

Stanford: EquivAct: SIM(3)-Equivariant Visuomotor Policies beyond Rigid Object Manipulation - Context ​

​

​The Perils & Promises of Fact-checking with Large Language Models (October 20th)

​

​Spoken Question Answering and Speech Continuation Using Spectrogram-Powered LLM (Revised Oct 20th)

​

To see New papers go here or here.

​

---

📱Semiconductor Notes📝

​

Apple: With its New M3 Chips, Apple joins the AI party​

​

​

​

🚀 Startups 🧠

​

Baichuan: Chinese AI start-up Baichuan claims to beat Anthropic, OpenAI with model that can process 350,000 Chinese characters​

​

​Shield AI raises $200M at a $2.7B valuation to scale military autonomous flying tech​

​

Anthropic: Dario Amodei’s prepared remarks from the AI Safety Summit on Anthropic’s Responsible Scaling Policy​

​

OpenAI: Exclusive: Ilya Sutskever, OpenAI’s chief scientist, on his hopes and fears for the future of AI​

​

---

​

💸Venture Capital🦄

​

​Former Google CEO Eric Schmidt Bets AI Will Shake Up Scientific Research (Bloomberg)

​

💼 Business to Business (B2B)📈

​

Kuo: Appleexpected to spend $4.75 billion on generative AI in 2024 alone​

​

🎬Videos🎥

​

[link] Dr. Fei-Fei Li, the 'Godmother' of AI, Talks Regulation

​

​

​

⚒️Novel Tools🔥

​

​The Always Up-to-date List of the 50 Best AI Tools in 2023 (ClickUp)

​

​Midjourney update brings ability to create custom styles​

​

👾Hacker News📰

​

Phind Claims Model beats GPT-4 at coding, with GPT-3.5-like speed and 16k context (Cannot be independently confirmed)

​

Meta’s Yann LeCun joins 70 others in calling for more openness in AI development (TechCrunch)

​

White House unveils AI.gov in a historic move towards comprehensive AI oversight​

​

OpenAI’s ChatGPT gets GPT-4 32K support for PDF and file chat with longer documents​

​

🤖 Robotics 🦿

​

(Repost) *Eureka: With GPT-4 overseeing training, robots can learn much faster - Nvidia Blog - Paper ​

​

​Electric Sheep Startup: Building a full-stack robotics company in the age of large AI models (The Robot Report)

​

🥼Healthcare🩺

​

​A year in review: AI’s evolving role in drug discovery and development in 2023​

​

New rules for health agencies that use AI (Politico) ​

​

​

🧪 Science 🔬

​

Alphabet: Google Deepmind shows the next generation of AlphaFold​

​

Meta:Toward a real-time decoding of images from brain activity​

​

​A method to interpret AI might not be so interpretable after all (MIT)

​

​

💬 Open-Source LLMs 👐

​

Hindi LLM: Introducing NOLIN​

​

​Zephyr: Direct Distillation of LM Alignment​

​

Anyscaleservice enables fine-tuning of open source LLMs (it’s called Endpoints)

​

💰Stock Market 🛒

​

Palantir shares pop 20% after its third-quarter results beat estimates on back of AI​

​

AMD Predicts $2 Billion in Sales of AI Chips, Boosting Market Confidence​

​

​

💀AGI and ASI 👽🌋

​

​This AI Threat Summit Is a Doomer’s Paradise (Gizmodo)

​

🎓 Education & Learning📚

​

​From Data Platform to ML Platform (Towards Data Science)

​

Finetuning LLMs with LoRA and QLoRA: Insights from Hundreds of Experiments​

​

​

➡️ Prompt Engineering 🧠

​

RAGis Just Fancier Prompt Engineering (AIM)

​

🎥Videos & Podcasts🎙️

Bloomberg Originals: Inside the Race to Build Safe AI | Exponentially with Azeem Azhar - [link]

​

​

​

​

🎖️National Defense & Military🕊️

​

​Shield AI raises $200M at a $2.7B valuation to scale military autonomous flying tech​

​

​Managing AI Risks in an Era of Rapid Progress (A short paper)

​

​

🔮Future of Work & Automation🦾

​

​The biggest winners — and losers — in the coming AI job apocalypse (Insider)

​

​

🐲 China A.I.🏯

​

​What is Alibaba's Tongyi Qianwen 2.0?​

​

Chinese scientists create chip that can perform AI task 3,000 times faster than Nvidia’s A100: study​

​

​

🦹🏻‍♂️Dystopian 🗽

​

​The Creepy New Digital Afterlife Industry ​

​

🏛Governance and Regulation⚖️

​

AI Safety Summit UK: 8 Hours Livestream (Day 1)

​

​Challenges in evaluating AI systems (Anthopric)

​

‘Existential to who?’ US VP Kamala Harris urges focus on near-term AI risks​

​

💬Opinion Editorial🙄

​

​Everything We Know About Humane’s Bewildering New AI Pin​

​

📣Public Relations💞

​

Microsoft, LinkedIn: LinkedIn passes 1 billion members and launches new AI chatbot to help you get a job​

​

🙉Last Memes🎭

Elon Musk repeating his warnings while making fun of the AI Summit he was attending with a sad meme.

There’s a general moral decline in leadership around technology.

A special message from Andrew Ng for Halloween.

​

🏋️‍♀️ Thanks for skimming, I hope you managed to find something of interest. 👀

​