Sam Altman has been very active on X the past few days, teasing the ChatGPT upgrades that are coming soon. He hyped the GPT-4o update that OpenAI delivered quietly over the weekend, which came just days after the CEO mentioned the roadmap ahead. ChatGPT will get GPT-4.5 in the coming weeks, with the big GPT-5 upgrade set to debut in the months that follow.
Altman also gave us a teaser about GPT-4.5 being a “feel the AGI” moment during testing, but I don’t know what to make of it. Just a month ago, the same CEO downplayed rumors about an imminent ChatGPT upgrade and the arrival of AGI, telling people to “pls chill and cut your expectations 100x!”
“Twitter hype is out of control again,” Altman tweeted on January 20th. “We are not gonna deploy AGI next month, nor have we built it. We have some very cool stuff for you but pls chill and cut your expectations 100x!”
These comments were made after the OpenAI CEO penned a blog post about the company knowing how to build AGI and preparing for superintelligence. Separately, rumors mentioned OpenAI’s plans to release AI Agents this year. Operator and Deep Research were released for ChatGPT after Altman’s tweet.
Fast-forward to February 17th, Altman is doing plenty of AGI hype on X on his own. “Trying GPT-4.5 has been much more of a ‘feel the AGI’ moment among high-taste testers than I expected!” he cryptically said, without revealing additional information.
This is how Twitter hype will get out of control again.
Of note, Altman penned an incredibly optimistic post about ChatGPT AGI and its effects on the world a few days ago to coincide with the AI Action Summit in Paris, where AI regulation was a big topic.
Here’s how Altman defined AGI at the time, complete with the fine print:
Systems that start to point to AGI* are coming into view, and so we think it’s important to understand the moment we are in. AGI is a weakly defined term, but generally speaking we mean it to be a system that can tackle increasingly complex problems, at human level, in many fields.
*By using the term AGI here, we aim to communicate clearly, and we do not intend to alter or interpret the definitions and processes that define our relationship with Microsoft. We fully expect to be partnered with Microsoft for the long term. This footnote seems silly, but on the other hand we know some journalists will try to get clicks by writing something silly so here we are pre-empting the silliness…
The OpenAI-Microsoft definition of AGI is AI that can generate at least $100 billion in profits. That’s what reports say, and that’s probably why Altman needed an asterisk in that post.
Put differently, AGI isn’t an objective concept, which means it might be difficult for humans to realize when AGI has been reached or whether they’re working with AGI rather than regular, dumber AI.
From the AGI point of view, all the advanced AIs we currently have fall into the dumber AI category despite their amazing abilities. These AIs are a massive resource for humans, but they’re not ready to tackle problems at a human level.
ChatGPT GPT-4.5 will surely get us closer to AGI. As a longtime ChatGPT user, I’m looking forward to the upgrades. But Altman’s hype is, by his own metrics, unwarranted. I doubt that we, regular ChatGPT users, will feel the AGI once GPT-4.5 drops. OpenAI engineers who train these models and have unrestricted access to them might feel it, sure.
So why is Altman suddenly open to revealing the ChatGPT roadmap, talk open-sourcing the AI, and hype AGI? Something happened between January 20th and February 17th.
Chinese startup DeepSeek released a reasoning model as good as ChatGPT o1 but much cheaper to train. This development shocked the US stock market, as investors worried AI hardware supremacy wouldn’t be enough to guarantee returns. Chinese engineers used software innovations to deal with the US ban on AI chips and develop DeepSeek R1 cheaper than what it should cost.
The DeepSeek shock occurred at a critical moment, too. The US announced the $500 billion Stargate AI infrastructure initiative to benefit OpenAI.
Also, OpenAI is in the process of raising more capital to the tune of tens of billions. This involves a big ChatGPT partnership with SoftBank in Japan, which has already committed to spending $3 billion a year on a joint initiative with OpenAI in Japan.
Interestingly, SoftBank CEO Masayoshi Son teased that AGI is coming sooner than he thought. At the time, I explained this early AGI arrival might mean little to regular ChatGPT users like you and me. Who knows when we’ll actually get AGI compared to enterprise customers?
Finally, Elon Musk offered to buy OpenAI for nearly $100 billion. XAI launched Grok 3 on Monday, complete with reasoning capabilities.
All this pressure might explain why Altman is suddenly so active on X, hyping ChatGPT products like the upcoming GPT-4.5 with AGI talk.
