---

OpenAI's New Deep Research Model and AI Cost Reductions

A professional programmer's take on the latest advancements in AI technology and cost reductions.

In a recent conversation with IT Minister Ashwini Vaishnaw, OpenAI CEO Sam Altman discussed the company's new deep research model, addressing the costs of AI models and the potential for cost reductions. These remarks come as Chinese firm DeepSeek has built a foundational model comparable to leading AI models at a fraction of the cost.

Small Language Models and Cost Decreases

Mr. Altman acknowledged the decreases in cost when training AI models, particularly when focusing on so-called small language models. These models, while less advanced and less expensive than their larger counterparts, still contribute significantly to the field of AI.

The Cost of Advanced AI Models

Despite the decreases in cost when training AI models, Mr. Altman admits that advanced models remain pricey:

"It’s still expensive to train advanced AI models. To stay at the frontier, costs will rise on an exponential curve, but the returns to increase in intelligence are exponential in terms of the economic and scientific value you can create.”

However, Mr. Altman suggested that the cost per unit of intelligence is dropping by tenfold annually, reflecting the rapidly diminishing costs of AI model training.

Focusing on Cost Reduction

Indian entrepreneurs and startups, according to Mr. Vaishnaw, are adamant about achieving the next level of cost reductions. He cited the Chandrayaan Mission's low cost as an example of their prowess:

"Of course, we are creating certain very good data sets which will be used for training the models in an Indian context in terms of language and the cultural and regional nuances within the country."

These data sets have the potential to further advance AI technology and reduce costs for models tailored to Indian languages and cultural nuances.

The Potential of OpenAI's New Deep Research Model

OpenAI's new deep research model seeks to make intellectually demanding tasks far more efficient, particularly in research fields. Altman explained:

"If you are a scientist trying to cure some disease, deep research is surely not going to cure that disease on its own. But if you can farm out the tasks that took you a lot of time but were lower value like literature reviews or equipment ordering instructions, maybe you can be twice as efficient."

By automating lower-value tasks, researchers can theoretically increase productivity and focus on higher-value tasks.

Conclusion

OpenAI's latest advancements in AI technology and cost reductions could revolutionize the field and create new opportunities for scientific research and startup growth. As the costs of AI models continue to drop, the potential for innovation and increased productivity will only continue to rise.

---

As a professional programmer, I'm excited about the future of AI and look forward to exploring these advancements in greater depth. Stay tuned for more detailed analyses of AI technology and its applications in the programming world.

---

Last updated on March 15, 2025