OpenAI CEO Sam Altman asked users what they’d like to see the AI research lab build or fix in 2024. Responses highlighted exciting new frontiers along with improvements to existing capabilities.
Unsurprisingly, many respondents requested artificial general intelligence (AGI) – AI able to perform any intellectual task a human can. However, as Altman noted, generalized AI remains on the more distant horizon and will require continued research and breakthroughs.
Another common request centered on GPT-5, the next iteration of OpenAI’s groundbreaking generative text models. Given the hype and capabilities jump from GPT-3 to GPT-4, users eagerly anticipate what GPT-5 might enable in terms of language understanding and content creation.
Better voice interaction modes for ChatGPT also emerged as a popular ask, with users desiring more naturalistic and conversational speech capabilities. Higher rate limits allowing more queries per day additionally ranked highly on wish lists.
Some users lobbied for enhancements to reasoning, wokeness controls, personalization, browsing abilities, and overall coherence in long-form responses. OpenAI’s models still exhibit certain limitations in logic, topic maintenance, and personality.
Interestingly, many users voiced interest in OpenAI releasing select models and code open source, enabling wider community experimentation and development. However, OpenAI balances openness with the need for funding to support expensive computational research.
In total, the requests underscore both inspiring new possibilities and current shortcomings. As Altman stated, OpenAI aims to deliver on many sought-after improvements in 2024 alongside other surprises. But restless users should manage expectations as some desires require foundational discoveries in AI safety and reasoning. Even genius AI doesn’t develop overnight!