Total Failure

GPT-5 Isn't Happening

In partnership with

Hello readers,

Welcome to the AI For All newsletter! Super Bowl ads for AI were underwhelming, the AI Action Summit in Paris was an unmitigated disaster, AI agents are facing skepticism, and GPT-5 isn’t happening. Is AI’s failure now complete? Let’s find out!

Total Failure

A sinking feeling

Today’s newsletter is about failure — the failure of AI, the media, politicians, and the Kansas City Chiefs, which brings us to our first story: Super Bowl commercials. Despite not really having anything new to offer, OpenAI, Google, Meta, and Salesforce thought that a Super Bowl ad would be a good use of money. Well, the results are in for the top-rated commercials. Of the 57 commercials, OpenAI ranked #53, Meta ranked #44, Google ranked #14 (slow clap), and Salesforce isn’t even on the list.

Unless you’re Helmuth von Moltke, you will agree that an honest depiction of war doesn’t paint a positive picture. Even heroism is too often pyrrhic. The art of information warfare is not disimilar to the art of war. Like Stonewall Jackson said: mystify, mislead, surprise. When people don’t know what to believe, they become paralyzed, allowing so much to slip by undetected. I forgot where I was going with this. Oh right, Elon Musk.

Musk is waging all-out war against OpenAI, offering $97.4 billion to buy the startup, which complicates OpenAI’s transition to a for-profit. Though Sam Altman informally declined the offer, OpenAI’s board is legally obligated to consider it. Musk called Altman a swindler, and Altman called Musk insecure. Both are right.

In a post on X, Sam Altman shared a roadmap for GPT-4.5 and GPT-5. Orion was originally intended to be GPT-5, but OpenAI struggled to make a model worthy of the name (or the hype). So, Orion is now just GPT-4.5, and GPT-5 is not a new, distinct model but an amalgamation of OpenAI’s current models (as I predicted). What this indicates is that LLMs and scaling have truly plateaued.

In other news, the AI Action Summit in Paris was a disaster. The UK and the US refused to commit to the “open, inclusive, and ethical” development of AI so as not to stifle AI progress. So, who is right? No one. The AI being addressed here is either mediocre or entirely fictitious. Despite this, France wants to invest €109 billion into AI, and the EU wants to invest €200 billion — for what exactly? More compute? Anyway, these summits are so premature as to be farcical. It honestly feels like our society is just larping the 20th century as we meme ourselves to death. History does repeat itself, but as Karl Marx noted, the first time as tragedy, the second time as farce.

The media has largely failed to inform the public and businesses of the myths of AI, herding them to inevitable disappointment. One such myth is the incoming workforce of AI agents. At this year’s CIO Network Summit, 61% of attendees said they were “experimenting” with agents, but 21% said they weren’t using them at all. A poll found that the primary concern was a “lack of reliability.” Enterprises are slow, yet AI companies are expecting 2025 to be the year AI agents accelerate. I have my doubts.

🔥 Rapid Fire

The future of presentations, powered by AI

Gamma’s AI creates beautiful presentations, websites, and more. No design or coding skills required. Try it free today.

🔍️ Industry Insights

“Interest in open source AI is growing as the performance of more open foundation models closes the gap to proprietary AI platforms. A recent survey of more than 700 technology leaders and senior developers across 41 countries by McKinsey, the Mozilla Foundation, and the Patrick J. McGovern Foundation provides the largest and most detailed analysis of how enterprises are thinking about and using open source AI.”

Source: McKinsey

💻️ Tools & Platforms

  • Talo → Real-time AI translator for video calls

  • Unstructured → Data ETL for your LLM

  • Elastic → The elastic search AI platform

  • SpotDraft → Intelligent contract management

  • Amplemarket → AI sales copilot for teams