Sam Altman claims knowing what questions to ask trumps raw intelligence as AI advances — Users struggle to realize Copilot and ChatGPT's full potential, owing to poor prompt engineering skills
OpenAI CEO says knowing which questions to ask will outshine raw intelligence as AI advances rapidly.

The world is changing exponentially with the rapid emergence and adoption of generative AI and its advanced tools. As you might have noticed, AI-powered search engines like OpenAI's ChatGPT Search are redefining the search landscape, presenting bite-size, often well-curated, and to-the-point responses to queries.
Despite being deemed an illegal monopoly in search, a former Google engineer says the company may have bigger fish to fry with OpenAI's "temporary prototype" search tool. Search isn't the only area majorly impacted by the AI bubble, OpenAI CEO Sam Altman recently claimed that AI might be slowly reducing the importance of raw intelligence.
While appearing in a recent episode of the ReThinking podcast, Altman told organizational psychologist Adam Grant (via CNBC):
“There will be a kind of ability we still really value, but it will not be raw, intellectual horsepower to the same degree. Figuring out what questions to ask will be more important than figuring out the answer.”
Grant seemingly shared Altman's sentiments, reiterating the importance of knowing how to ask questions that seek to establish a pattern and establish more context on a topic:
"I think it’s much more valuable to be a connector of dots ... If you can synthesize and recognize patterns, you have an edge.”
Interestingly, the OpenAI executive acknowledges the importance of learning how to ask important questions that provoke a thought process that could potentially help solve complex issues. Last year, a report revealed that the top complaint in Microsoft's AI division — Copilot doesn't work as well as ChatGPT.
Get the Windows Central Newsletter
All the latest news, reviews, and guides for Windows and Xbox diehards.
I have certainly gotten the greatest professional joy from having to really creatively reason through a problem and figure out an answer that no one’s figured out before. What I expect to happen in reality is, there’s going to be a new way we work on the hard problems.
OpenAI CEO, Sam Altman
Microsoft quickly refuted the claim, citing a lack of proper prompt engineering skills. A Microsoft employee indicated that the quality of Copilot's response depends on how you present your prompt or query. Microsoft has since furnished Copilot users with a library of resourceful videos to help users enhance their prompt engineering skills.
In September 2024, Microsoft launched Copilot Academy to help equip businesses with the best practices when interacting and leveraging the tool's capabilities. That said, prompt engineering is increasingly becoming a big deal and an essential skill as more people hop onto the AI bandwagon and become more dependent on these tools.
Kevin Okemwa is a seasoned tech journalist based in Nairobi, Kenya with lots of experience covering the latest trends and developments in the industry at Windows Central. With a passion for innovation and a keen eye for detail, he has written for leading publications such as OnMSFT, MakeUseOf, and Windows Report, providing insightful analysis and breaking news on everything revolving around the Microsoft ecosystem. You'll also catch him occasionally contributing at iMore about Apple and AI. While AFK and not busy following the ever-emerging trends in tech, you can find him exploring the world or listening to music.
-
GraniteStateColin One of my favorite lines of all time in movies or TV, is said by Doctor Who, known for both his intelligence and knowledge, "The answers are easy. It's asking the right questions that's hard." (Doctor #4, Tom Baker, if anyone's curious.)Reply
I find the use of the word "intelligence" strange in this article. I would say that knowledge != intelligence. Knowledge is accumulated facts and learned skills. It's mostly a function of training and, for a human, a person's desire to learn and ability to remember. Intelligence is the ability to see or create new things (solutions, jokes, music, military strategy, physics theories, faking out the defense in a new play to get the ball to the basket,...) that didn't previously exist. So, posing the proper queries to AI to get at the knowledge it has would be a form of intelligence.
The AI can have vast knowledge. At the limit, it can equal the accumulated recorded knowledge of all humanity. AI is also good at finding patterns, which is a form of intelligence, but only one. This has been incredibly useful, for example, in predicting protein folding and chemical engineering for materials sciences leapfrogging attempts to actually model molecular interaction by electric charges -- a task like trying to exactly predict the weather months into the future, impossible due to the number of variables and chaotic interactions. AI is able to bypass this using pattern recognition at a scale that the human brain can't do. It's also helping with some cosmological questions where there are slight statistical variations only visible looking at thousands of data points, again beyond human capacity. These are wonderful but are no substitute for the human ability to go BEYOND patterns to imagine things and solutions that have no precedent.
At this point, AI is still just a tool that requires clever humans to ask it the insightful questions, just like the Doctor said. Maybe one day it will go beyond that, but if so, it will be different from the large language models we see today.