What Did OpenAI Do This Week? [FREE ISSUE]
An extremely busy week for OpenAI. News count: +20
NEW MODELS, A TIGHTER GRIP, AND A LOT OF LEAKS
Latest OpenAI Articles
OpenAI Preparedness Framework 2.0
Right before releasing o3, OpenAI updated its Preparedness Framework to 2.0.
Melon Usk - e/uto
•
3d
Yep, there are ways to align OpenAI and others from the outside by motivating people to bring GPUs into safe clouds (that will have an App Store for AI models), can be done profitably especially with gamers. I wrote about it not long ago
Askwho Casts AI
•
3d
Podcast episode for this post:
https://open.substack.com/pub/dwatvpodcast/p/openai-preparedness-framework-20
OpenAI #13: Altman at TED and OpenAI Cutting Corners on Safety Testing
Three big OpenAI news items this week were the FT article describing the cutting of corners on safety testing, the OpenAI former employee amicus brief, and Altman’s very good TED Interview.
"continue to think this ‘AGI is fungible’ claim is rather bonkers crazy."
This is an interesting issue because when I talk with investors and engineers in the AI space I feel like this idea is gaining ground. The idea that the LLMs are turning out to be commoditized, that for many or most AI things you build, the individual LLM you build with can be swapped out. There's some price-performance tradeoff and there's some friction in swapping but more or less it's a fungible component of your system.
The part that is *not* fungible is your user base. You can tune your product to do well on the specific tasks that your users do, and your competitors can't just match that because they don't have the specific data from your users, they can't launch experiments and get feedback from your user base.
The LLMs can still be very profitable; AWS and Azure both make a lot of money despite being generally fungible.
I'm curious why you think this line of reasoning is bonkers crazy. To me it seems at least a reasonably possible outcome. Do you think there will be a phase shift at some point, where all of a sudden you don't have many similarly capable LLMs, where one of them pulls far ahead?
Testing a model before it's released seems important. But it seems much less important than thoroughly understanding emergent dangers of powerful models in depth throughout the process of development, release, and operation. I worry that conflating safety with 'how much time did you spend testing before release' plays into arms race dynamics by putting up a fairly superficial barrier at the end of the model creation process. We don't want to suggest that creating ultra-powerful dangerous models internally is fine as long as what you release has enough safeguards that it doesn't do anything bad in the hands of users. That just accelerates the race and hides the biggest risks from the public.
Microsoft Breaks Free From OpenAI
Microsoft is betting on model independence—by backing Grok, it’s signaling a future where OpenAI is optional, not essential.
Good morning AI entrepreneurs & enthusiasts,
OpenAI AI Agent Design Guidance
Designing Safer AI Agents: The Power of Human Intervention
Thanks for reading Cobus Greyling on LLMs, NLU, NLP, chatbots & voicebots! Subscribe for free to receive new posts and support my work.
AI Free Resources By OpenAI and Google
This page compiles all the free resources offered by OpenAI and Google
It’s quite interesting to see so many resources available for free from many companies. I was looking for LLM agents and stumbled on free AI book offered by Open AI. Its a short book but offers good introductory information on the LLM topics.
🕵️ OpenAI Releases Its Image Generation API
⏩ How to Generate GIFs Directly Inside ChatGPT...
Hey AI fam! 👋
⚡️GPT 4.1: The New OpenAI Workhorse
Michelle Pokrass returns, with Josh McGrath to talk about the new GPT 4.1 model
We’ll keep this brief because we’re on a tight turnaround: GPT 4.1, previously known as the Quasar and Optimus models, is now live as the natural update for 4o/4o-mini (and the research preview of GP…
In the Matter of OpenAI vs LangGraph
The silent war in Agent Engineering gets loud.
Quick reminder: AI Engineer CFPs close soon! Take a look at “undervalued tracks” like Computer Use, Voice, and Reasoning, and apply via our CFP MCP (talks OR workshops, we’ll figure it out).
Thanks for this!
The terminology of 'chain' in LangChain lingo has never been clear to me since it seems to refer to both individual nodes and the entire workflow, so I'm happy it has been superseded by 'workflow'. LangGraph is their best to date imo.
Anthropic's Building Effective Agents article makes the distinction between agentic loops and deterministic workflows clearer than any other thoughtpiece, even though it oversimplifies the distinction. I get the impression that most people building in this space (myself included) see it more as a spectrum of agency.
I'm keen to see this Agent Framework Breakdown evolve. It's a great start for what could ultimately become a very useful resource.
125: Visa, PayPal, OpenAI kill the checkout
Mastercard brings stablecoins to 150M stores. SaaS is now a national security risk. Visa, Mastercard & PayPal roll out AI agent payment rails. BCG’s point of view on AI agents. And more.
Hey, it’s Marc.
Boston Dynamics led a robot revolution+AI
Microsoft and OpenAI drifting apart
Boston Dynamics led a robot revolution, now with AI
The IRS Leaves Crypto Alone!
Will Tether Overcome OpenAI?
The global cryptocurrency market cap today is $3.04 trillion, a 2.9% change in the last 24 hours. Total cryptocurrency trading volume in the last day is at $60.5 billion.
OpenAI Launches GPT-Image-1 API
OpenAI launches a powerful image model API, Anthropic warns AI employees may arrive soon, and Instagram drops Edits—a new app for pro-level video creation made easy.
Imagine this: You're designing a stunning image for your brand, powered by AI that understands style, context, and even your brand voice. That’s now a reality—thanks to OpenAI’s new image model, which just launched in their API. Big names like Adobe and Figma are already using it to supercharge creative workflows, and now, you can too.
I Read Everything From OpenAI About Building Agents
So you don't have to.
One thing I learned from John Carmack is that “you should read the docs”:
#9 - Build Your First AI Agent
OpenAI Can Teach You How 🔐
Welcome to the 9th issue of AI Agents Simplified! 🎉
Breaking: Top OpenAI Catastrophic Risk Official Steps Down Abruptly
It's the latest shakeup to the company's safety efforts
OpenAI's top safety staffer responsible for mitigating catastrophic risks quietly stepped down from the role weeks ago, according to a LinkedIn announcement posted yesterday.
Not releasing a safety report for this new, clearly frontier, model completely goes against OpenAIs prior commitments, and makes me incredibly pessimistic for the future of AI safety. There’s a new scandal with OpenAI every few weeks, but you don’t hear many from other labs. I wonder if it’s particular to OpenAI, if it’s Sam Altman, or if the other companies are just better at hiding it. I hope it’s just an OpenAI problem and safety measures at other labs are moving fast and can eventually be copied by OpenAI and other labs. Fingers crossed for you mech interp researchers out there 🤞
Everyone should be alarmed by what’s happening at frontier AI labs. These companies aren’t just tech firms—they’re now functioning as de facto defense contractors, and their machine learning engineers are, in effect, weapons researchers. The sudden departure of key safety personnel only underscores how unstable and opaque these institutions have become.
This isn’t just a U.S. issue. The global implications are massive. For example: Canada is currently in the middle of an election, and not one major party is addressing the immediate economic threat of AI-driven workforce collapse or the strategic risk posed by nationalist artificial superintelligence developed by U.S. firms under the Trump administration.
When the stakes are this high and the silence is this loud, the danger isn’t speculative—it’s systemic.
I wrote about this in more detail here:
The Real Threats Your Leaders Won’t Acknowledge (Until It’s Too Late)
OpenAI AI Agents SDK Guardrails
Well-designed guardrails assist in managing data privacy risks. For example, preventing system prompt leaks…
Thanks for reading Cobus Greyling on LLMs, NLU, NLP, chatbots & voicebots! Subscribe for free to receive new posts and support my work.
IGN vs OpenAI | AI and Games Newsletter 30/04/25
Plus conference videos, a new case study, an Alien teaser and more!
The AI and Games Newsletter brings concise and informative discussion on artificial intelligence for video games each and every week. Plus summarising all of our content released across various chann…
AI-Powered Technical Analysis in *Google Sheets* (for free)
Just Bring Your Own OpenAI API key (BYOK)
AI for Charts (public beta) is a free Google Sheets Add-On that connects your Google Sheets spreadsheets directly to the OpenAI Developer API with your own OpenAI API key.
Build Custom Financial Datasets with AI Agents (without Web Scraping or Finance APIs)
OpenAI Agents SDK with OpenAI or Perplexity Sonar Models
This tutorial shows you how to build a custom AI-driven workflow to automatically collect and structure financial datasets without relying on traditional web scraping or expensive financial APIs. You'll learn how to create AI agents with OpenAI's Agents SDK Python library, output structured data using the Pydantic library that is ready for additional LL…
Who is Reid Hoffman?
LinkedIn CoFounder, Microsoft & OpenAI board member & Bipartisan advocate
Key Themes and Important Ideas/Facts:
Sam is Teaching us How to Build AI Agents!
OpenAI - A Practical Guide to Building Agents ⚡
We’ve been talking about AI Agents for weeks! how they’re changing workflows, how to build them, how they’re not just smart chatbots... but now?
Nice work Sam
But here is another guide that you will find useful
Great article
Edition 210: Ali Rohde Jobs
Roles at Mercury, OpenAI, Plaid, Ramp, Sesame AI, CZI
Welcome to Edition 210 of Ali Rohde Jobs! As always, if you have a relevant job you’d want to share, let me know — Ali
AI: Microsoft & OpenAI, it's complicated. RTZ #705
...more misalignment on an iconic AI partnership
The most iconic partnership of this AI Tech Wave, Microsoft & OpenAI, is showing more signs of stress. The Tango is seeing some missed steps together.
Related OpenAI Substacks
