📈 How are companies using AI?

➜ Apple quietly bought Canada’s DarwinAI startup. Bloomberg suggests that this purchase may be related to the promise made by Tim Cook to announce something big in AI this year.

Bloomberg

➜ It's interesting to see how pricing strategies for AI evolve. With its new Copilot for Security (an AI-powered chatbot for cybersecurity professionals), Microsoft will charge businesses $4 per hour as part of a new consumption model.

The Verge

➜ A new survey shows a mismatch between CEOs and their teams when it comes to AI expectations. Most CEOs prioritize cutting costs and building customer loyalty as the main drivers for AI adoption.

Forbes

➜ By 2028, more than 50% of enterprises that have built their own LLMs from scratch will abandon their efforts due to costs, complexity, and technical debt, claims Gartner.
Gartner

➜ Oracle is finally adding GenAI features to handle specific tasks in its finance and supply chain software, for example, summarizing a long chain of back-and-forth price negotiations with a supplier.

Reuters

➜ Andreessen Horowitz published new data on the Top 100 GenAI Apps and Websites. Striking stat: over 40 percent of the companies have not been on the list six months ago.

a16z

💡 What are the nerds up to?

➜ If you can understand a spreadsheet, then you can understand AI. Here’s a video course that uses a real LLM implemented entirely in Excel.

Spreadsheets are all you need.ai

➜ First AI software engineer? Devin, an autonomous agent, passed practical engineering interviews from leading AI companies and completed real jobs on Upwork. Performance: Devin correctly resolves 13.86% of the issues unassisted (the previous score was below 2%).

Cognition Labs

➜ 'Apple's so back!' – says Jim Fan, NVIDIA research manager. His excitement was due to Apple dropping a paper on their MM1 (multimodal foundation model), revealing a surprising amount of technical details. It seems they also train on GPT-4V-generated data.

Jim Fan 

➜ Do you pause to think before you speak? It turns out that language models can learn to reason in a similar manner. A group of researchers introduced Quiet-STaR, a generalization of STaR in which LMs learn to generate rationales at each token to explain the future text, improving their predictions.

Cornell University

➜ AI makes us more human, claims OpenAI’s Head of ChatGPT Peter Deng. Dive into a 1-hour-long conversation from SXSW 2024. Worth it.

YouTube

Thanks for reading!

I share these stories because I believe it's important for all of us to keep up with AI. To support my mission, share this newsletter on LinkedIn.