Following OpenAI’s eventful 12 days of shipmas extravaganza, the company CEO, Sam Altman, had an open discussion on X, asking users, “What would you like OpenAI to build/fix in 2025?” As expected, the post went viral, with a multitude of users blatantly expressing their AI wants and needs.
Right off the bat, a user expressed the need for a drag-and-drop user interface featuring multiple separate models that can be strung together to generate and refine outputs. The discussion was previously featured in OpenAI’s Developer Forum, where users indicated, “I think you will get better results if you divide the task among multiple models, each one optimized to do one task very well.”
The executive referred to the user’s submission as “an interesting idea” that he’d not considered. Another user shared a comprehensive list of requests, including a stronger GPT-4o replacement, GPT/o/Sora seamlessly integrated into ChatGPT, longer context, aggressive updates, and a “middle ground” paid subscription plan costing approximately 50-70 dollars. While cryptic, OpenAI CEO Sam Altman indicated that the user would be quite happy with the company’s progress this year.
Interestingly, another user requested the executive loosen the guardrails on its next-gen AI models, claiming it would be “cleaner.” Sam Altman indicated there’s a need for a “grown up mode.”
Family accounts. Let me create accounts for my kids with guard rails. Let their curiosity take off, but within reasonable limits, as determined by the parent. Maybe we could even get insights about our kids from their usage!December 25, 2024
At the same time, another user requested family accounts, which would allow them to create accounts for children with guardrails. The user indicated that this approach would let their curiosity take off. However, the user stated that it was paramount to have limits established by the parents to prevent the system from spiraling out of control.
Finally, @davidlach on X requested proper agents, prompting Sam Altman to indicate “Happy 2025.” As you may know, OpenAI is reportedly gearing up to ship an AI agent dubbed Operator in January 2025. Late last year, Chief Product Officer Kevin Weil indicated:
“I think 2025 is going to be the year that agentic systems finally hit the mainstream.”
What to expect from OpenAI in 2025 and beyond
The thread on Sam Altman’s post on X asking users what they’d like OpenAI to build or fix in 2025 is seemingly endless, but the executive might have hinted at several updates and benchmarks we could look forward to in 2025.
OpenAI could release a GPT-4o successor sporting better capabilities. The ChatGPT maker’s advanced AI models may feature better memory with longer context when generating prompts. We could also get a “grown up mode” with fewer guardrails, though there are safety and privacy concerns riddling its development. Finally, Sora might get a significant update, scaling its capabilities to greater heights and potentially longer videos.
common themes:AGIagentsmuch better 4o upgrademuch better memorylonger context“grown up mode”deep research featurebetter soramore personalization(interestingly, many great updates we have coming were mentioned not at all or very little!) https://t.co/lMZmlZif66December 30, 2024
Of course, AGI was at the top of Sam Altman’s list despite rumors that the ChatGPT maker might have already hit the coveted benchmark after releasing its o1 reasoning model to broad availability. Interestingly, around the same time, a separate report suggested that OpenAI is planning to scrap a stringent AGI clause that could sever its partnership with Microsoft to secure future funding and access to cloud computing.
Based on the highlighted information, there’s a possibility that OpenAI may have already hit the coveted benchmark or, at the very least, could be on the verge. Interestingly, Sam Altman indicated that AGI is achievable with current hardware within the next 5 years and would whoosh by with surprisingly little societal impact.
That said, OpenAI could have more surprises in store for users in 2025. As Sam Altman indicated, “Many great updates we have coming were mentioned not at all or very little!”
