After OpenAI captivated tech enthusiasts with its electrifying 12 days of shipmas extravaganza, Sam Altman, the company’s CEO, took to X with an intriguing question: “What would you like OpenAI to build or fix in 2025?” This sparked quite a buzz as users flocked to voice their desires and suggestions for the future of AI.
Kicking things off, one tech-savvy user called for a drag-and-drop interface featuring multiple models that could be interconnected to craft and refine outputs. Previously spotlighted in OpenAI’s Developer Forum, this idea resonated with many who believe that splitting tasks among specialized models could yield superior results. Altman himself found the proposal “interesting,” admitting he hadn’t considered it before.
As the dialogue unfolded, another user laid out a wishlist that included a robust GPT-4o successor, seamless integration of GPT/o/Sora into ChatGPT, extended context capabilities, more frequent updates, and a new mid-tier subscription plan priced between $50 to $70. Even though some aspects were vague, Altman hinted that the user might find themselves quite pleased with what OpenAI has planned for the year.
In a twist, someone else suggested easing the restrictions on future AI models, labeling this potential change as “cleaner.” Altman acknowledged the call for a “grown-up mode,” suggesting a flexible approach to the model’s functionality.
A user also pitched the idea of family accounts, envisioning a feature where parents could create child-friendly profiles with appropriate controls woven in. This setup would allow kids to explore while keeping their curiosity contained within parental guidelines. The notion of gaining insights about children from how they use such accounts added an intriguing layer to the proposal.
Meanwhile, as @davidlach on X requested refined AI agents, Altman cheerfully responded with “Happy 2025,” hinting at exciting developments on the horizon. OpenAI is rumored to be preparing for the launch of “Operator,” an AI agent, by January 2025. Just last year, Product Chief Kevin Weil suggested that 2025 could be the breakthrough year when agentic systems truly gain traction.
Looking forward, there are whispers about OpenAI releasing a new and improved GPT-4o model with enhanced capabilities. We might see ChatGPT’s AI models equipped with better memory and extended context generation. The “grown-up mode” is on the table, though it brings with it substantial considerations regarding safety and privacy. Significant updates for Sora are also anticipated, potentially enhancing its functionality and video capabilities tremendously.
Discussions about AGI (Artificial General Intelligence) continue to swirl around Altman’s future vision, arousing speculation about whether OpenAI has already reached significant milestones. Talks even suggest that OpenAI could reconsider its AGI clause with Microsoft to open new avenues for funding and cloud resources. Altman has hinted that achieving AGI could be within reach with current technology in the next five years, sparking debates on its potential societal impacts.
As always, OpenAI might have a few more surprises up its sleeve as 2025 unfolds. Altman himself teased, “Many great updates we have coming were mentioned not at all or very little!” There’s certainly much to look forward to in the evolving world of AI.
