🥟 Chao-Down #337 Former OpenAI employees pen open letter on 'serious risks' of AI, Apple's deal with Sam Altman to catch up on AI, Google scales back its AI overviews following backlash
Plus, GenZ and millennials are ready for AI to manage their money.
Is the AI revolution already losing steam? That’s the question posed by the Wall Street Journal highlighting how tech companies are finding it hard to justify the ROI on the massive spend and costs they’ve incurred in trying to cement their leadership in the space.
From the piece:
The rate of improvement for AIs is slowing, and there appear to be fewer applications than originally imagined for even the most capable of them. It is wildly expensive to build and run AI. New, competing AI models are popping up constantly, but it takes a long time for them to have a meaningful impact on how most people actually work
To me, we’re in the phase where we’re spending more on building AI than figuring out what we can meaningfully do with it. For us to realize AI’s potential, we’ll need more than just incremental productivity gains. AI needs to enable new markets, products, and experiences that were previously impossible before.
NVIDIA and the hardware makers have claimed all the spotlight now. Who else will capture the value and meet the eye-watering expectations we’ve put on AI?
-Alex, your resident Chaos Coordinator.
What happened in AI? 📰
Former OpenAI employees say AI companies pose 'serious risks' (qz.com)
Apple Made Once-Unlikely Deal With Sam Altman to Catch Up in AI (MSN)
Five ways criminals are using AI (MIT Technology Review)
Gen Z and millennials are ready for AI to help manage money (qz.com)
Google seemingly scaled back AI Overviews after its rocky launch (androidpolice.com)
AI training data has a price tag that only Big Tech can afford (TechCrunch)
Always be Learnin’ 📕 📖
LLM RAG Paradigms: Naive RAG, Advanced RAG & Modular RAG (Medium)
Hi, AI: Our Thesis on AI Voice Agents (a16z.com)
Implementing FrugalGPT: Reducing LLM Costs & Improving Performance (portkey.ai)
Projects to Keep an Eye On 🛠
ragapp/ragapp: The easiest way to use Agentic RAG in any enterprise (Github)
databonsai/databonsai: Clean & curate your data with LLMs. (Github)
rjmacarthy/twinny: The most no-nonsense, locally or API-hosted AI code completion plugin for Visual Studio Code - like GitHub Copilot but completely free and 100% private. (Github)
The Latest in AI Research 💡
Let's Think Dot by Dot: Hidden Computation in Transformer Language Models (arxiv)
Evolutionary Optimization of Model Merging Recipes (arxiv)
GRAG: Graph Retrieval-Augmented Generation (arxiv)
The World Outside of AI 🌎
What we know (and don’t know) about Ozempic (Vox)
How DeviantArt died: A.I. and greed turned a once-thriving community into a ghost town. (Slate)
God Inc.—Church Startups Spread Franchise Model Across U.S. (WSJ)
Can mathematicians help to solve social-justice problems? (Nature)
In-person friendships are better for health than virtual connections (The Washington Post)
Your dreams can have meaning, scientists say (NPR)
One Last Bite 😋
