In a bold statement on the future of artificial intelligence, OpenAI CEO Sam Altman asserted that AI performance will continue to improve with increasing investment, while the cost of accessing AI will drastically decline. In a blog post on Sunday, February 9, Altman claimed that AI advancements will remain tied to rising expenditure, emphasizing that training compute, data, and inference compute are the primary drivers of intelligence in AI models.
AI Costs to Plummet, Widespread Adoption Expected
Altman highlighted a 10x annual drop in AI costs, stating that this decline will fuel mass adoption and widespread integration of AI systems. He cited OpenAI’s GPT-4 to GPT-4o transition as an example, where token costs fell by 150x between early 2023 and mid-2024.
His remarks precede the Paris AI Action Summit on February 10-11, where he is expected to discuss AI economics and governance with global leaders and tech executives.
OpenAI’s Expansion and Funding Push
OpenAI is reportedly in talks to raise $40 billion in fresh funding, aligning with broader industry efforts to build AI infrastructure. Tech giants including Microsoft, Nvidia, Oracle, and SoftBank have pledged $500 billion towards data centers, power grids, and computing facilities in the United States over the next four years.
Competition With Open-Source AI Models
Altman addressed the ongoing debate between open-source and closed AI models, hinting at OpenAI’s need for a revised open-source strategy following the rise of DeepSeek, a Chinese AI startup that recently introduced a low-cost AI model, rattling the global tech industry.
“Many of us expect to give people more control over AI, including open-sourcing more,” Altman said, acknowledging the trade-offs between safety and individual empowerment in AI development.
AGI and Its Economic Impact
Discussing Artificial General Intelligence (AGI), Altman reinforced his belief that AGI-level systems are approaching reality, envisioning a future where millions of AI agents assist in knowledge work across industries. However, he admitted that AGI systems will still lack groundbreaking creativity and have unexpected weaknesses.
Altman warned of the disruptive economic impact of AGI, predicting that the cost of many goods will plummet, while luxury items and scarce resources like land will become even more expensive. He also raised concerns about AI being misused for mass surveillance by authoritarian regimes.
Clarification on OpenAI’s Microsoft Partnership
In a footnote, Altman stressed that his statements on AGI do not impact OpenAI’s relationship with Microsoft. The companies reportedly define AGI as an AI system that generates at least $100 billion in profits, and Microsoft has invested over $13 billion in OpenAI to date.
“We fully expect to be partnered with Microsoft for the long term,” Altman reaffirmed.
The Road Ahead
As AI evolves rapidly, Altman’s insights reflect both tremendous opportunities and looming challenges in the field. With AI costs set to decline and AGI on the horizon, the coming years could witness an unprecedented transformation across industries, economies, and governance frameworks worldwide.