- AI For All
- Posts
- The Weight of Expectations
The Weight of Expectations
PLUS: Figure 02
Hello readers,
Welcome to another edition of This Week in the Future! OpenAI is lowering expectations for this year’s DevDay and experienced more high profile departures, Figure unveiled its next-generation Figure 02 humanoid robot, and NVIDIA’s new flagship AI chips have been delayed. What does all this mean for the AI space? Let’s find out!
The Weight of Expectations
OpenAI continues to be an enigma. One that has captured people’s attention for so long that they’re hoping there is something at the end of this that makes it all worth it. The weight of expectations, or weights in AI’s case 😆, has never been so heavy. However, OpenAI is tempering expectations for this year’s DevDay by saying, “While we know developers are waiting for our next big model, which we shared has begun training earlier this year, these events will focus on advancements in the API and our dev tools.”
OpenAI did quietly release GPT-4o-2024-08-06, a slightly updated model now available on Azure. OpenAI CEO Sam Altman also tweeted a picture of strawberries, leading many to conclude that Sam Altman likes strawberries … and also that the reveal of the secretive Project Strawberry (which we covered here) is imminent.
Meanwhile, OpenAI experienced yet another leadership exodus. President and co-founder Greg Brockman is on extended leave until the end of the year, co-founder John Schulman jumped ship to Anthropic, and product manager Peter Deng also left some time ago. Some have asked why this would happen if a breakthrough were imminent.
Very much dependent on OpenAI making progress is Figure, a robotics company that uses OpenAI’s vision models for its humanoid Figure 02. As far as robotics progress, Figure 02 has a sleeker design and 16 degrees of freedom in its hands, but it’s still infuriatingly slow. Maybe DeepMind’s ping-pong robot can help them out.
To top all this off, NVIDIA, who very much everyone is dependent on (and who Elliott Management issued a dire warning about), has been forced to delay its new flagship Blackwell chips until 2025 because of a design flaw.
Our Take
OpenAI is like a TV show. We’re at the point where audiences have devoted years of investment into the show, and now it’s time for some kind of grand finale. As long as the finale is yet to arrive, people can remain in denial about the cracks forming. “It’s all leading to something!” And then nothing. Betrayal turns into resentment, and you end up wishing the show you loved had never existed, so that you would be spared your pain. Or, you know, products just take some time. 😐️
🔥 Rapid Fire
OpenAI introduces structured output for developers in API
Mistral AI unveils model customization, Agents, and SDK
Meta announces call for applications for Llama 3.1 Impact Grants
NVIDIA scrapes human lifetime of videos per day to train AI model
Google stealth acquires Character AI and rehires founders
OpenAI won’t yet release ChatGPT detector with 99% accuracy
Elon Musk sues OpenAI and Sam Altman again over mission
Google announces addition of Gemini to smart home products
LG AI Research showcases EXAONE 3.0, an open source LLM
Alibaba releases Qwen2-Math, outperforming GPT-4 at math
IBM introduces generative AI-powered cybersecurity assistant
Palantir and Microsoft partner to deliver AI for national security
S&P Global and Accenture partner to train 35,000 employees on AI
Microsoft scales back Emirati AI partnership over China links
Britain cancels $1.7 billion in computing projects for AI
UK launches investigation into Amazon and Anthropic partnership
Google pulls controversial Olympics advertisement for Gemini
Humane’s panned Ai Pin sees daily returns outpace sales
AI detects heart attack risk, hailed as ‘game changer’ by scientists
AI predicts 3D structures of receptors for drug development
University of Chicago charts current ChatGPT adoption
AI researcher François Chollet says LLMs are a dead end to AGI
Verizon uses AI and machine learning to prevent fiber cuts
The fastest way to build AI apps
Writer is the full-stack generative AI platform for enterprises. Quickly and easily build and deploy AI apps with Writer AI Studio, a suite of developer tools fully integrated with our LLMs, graph-based RAG, AI guardrails, and more.
Use Writer Framework to build Python AI apps with drag-and-drop UI creation, our API and SDKs to integrate AI into your existing codebase, or intuitive no-code tools for business users.
📖 What We’re Reading
“Mentions of AI in retailers’ earnings calls soared last year—which was no surprise, given that gen AI is poised to unlock between $240 billion to $390 billion in economic value for retailers, equivalent to a margin increase across the industry of 1.2 to 1.9 percentage points. This, combined with non-generative AI and analytics, could turn billions of dollars in value into trillions.”