OpenAI’s new AI model shift & Amazon takes on Nvidia

OpenAI’s smarter approach to scaling AI, Apple’s quirky notification AI, and Amazon’s battle with Nvidia for AI chip dominance.

November 12, 2024

An AI-powered professional will earn 10x more. 💰

An AI-powered founder will build & scale his company 10x faster 🚀

An AI-first company will grow 50x more! 📊

🚀Join this 3-hour AI Workshop (worth $399) - FREE for Onyx Data readers to learn AI strategies & hacks to 10X work output and grow your business.

🗓️ Tomorrow | ⏱️ 10 AM EST

With AI & Chatgpt, you will be able to:

✅ Make smarter decisions based on data in seconds using AI

✅ Automate daily tasks and increase productivity & creativity

✅ Skyrocket your business growth by leveraging the power of AI

✅ Save 1000s of dollars by using ChatGPT to simplify complex problems

Latest headlines 👇

As AI scaling strategies hit roadblocks, OpenAI and others are shifting to methods that make algorithms more efficient and human-like in their reasoning. Instead of focusing solely on expanding data and computing power, OpenAI’s new “o1” model uses a technique called “test-time compute,” allowing models to assess multiple answers during use to make better decisions. This shift could alter AI's demand for costly hardware like Nvidia's chips, as AI labs, including Google and Anthropic, develop similar methods. Read more

Apple’s latest AI-powered feature on iPhones, iPads, and Macs summarizes app notifications, aiming to simplify updates from chats, emails, and home devices. Sometimes, it’s spot-on—like summarizing “Garage recently closed” instead of a long list of updates—but often, the mini-AI’s attempts are more amusing than helpful. From relationship breakups to “10 people detected at front entrance,” its summaries can be vague, missing context, or unintentionally funny. While it’s no game-changer, Apple’s notification AI at least keeps things interesting. Read more

Elon Musk might sway Donald Trump’s administration to adopt stricter AI safety regulations, according to MIT’s Max Tegmark. Musk’s support for AI oversight, including a recent California bill vetoed by the governor, shows his continued concern about unrestrained AI development. Tegmark believes Musk could help Trump recognize the dangers of competing to create AGI (artificial general intelligence), likening it to a “suicide race.” Trump has indicated he may repeal Biden-era AI safety measures, which Musk and Tegmark support to ensure responsible AI growth. Read more

Amazon is intensifying efforts to challenge Nvidia in the AI chip market, unveiling its latest Trainium 2 chips through Annapurna Labs. Aimed at reducing reliance on Nvidia, these custom chips are tested by AI players like Anthropic and offer 40% cost savings in generating AI responses. Amazon is investing heavily, expecting $75 billion in tech infrastructure spending in 2024. Though Nvidia still dominates with $26.3 billion in recent revenue, Amazon aims to offer clients more cost-effective options and reduce dependency on Nvidia’s near-monopoly. Read more

A new Slack study reveals that U.S. workers’ AI usage has plateaued, with only 33% now using AI at work — a mere 1% increase from previous months. This slowdown in adoption could challenge optimistic market projections for AI. Globally, adoption grew slightly, reaching 36%, though nearly half of workers feel hesitant to disclose AI usage to managers, fearing perceptions of laziness or incompetence. Slack urges leaders to better train and openly support AI experimentation to boost comfort and adoption. Read more

Adobe has reaffirmed that it won’t use customer data to train its AI models, addressing concerns from creatives who rely on its software. The company’s generative AI, branded as Firefly, only uses content Adobe has rights to, making it "commercially safe." A recent terms of service update had sparked worry, but Adobe clarified that customer data remains off-limits for training. However, Adobe does use contributions from Adobe Stock for AI training, offering bonuses to those contributors. Read more

OpenAI recently purchased the Chat.com domain from HubSpot co-founder Dharmesh Shah for $15 million. Shah initially acquired the domain to build an app similar to ChatGPT, assuming OpenAI would remain a platform-only company. Once he saw OpenAI’s commitment to ChatGPT, Shah decided to sell the domain, admitting he wasn’t interested in competing with them. Read more