Microsoft CEO Satya Nadella recently spoke to Wired’s Steven Levy about the company’s significant investment in generative AI and its pervasive integration into consumer and enterprise products. Nadella acknowledged his awe at the potential of AI technology, particularly during the transition from GPT 2.5 to 3, when significant achievements were made without explicitly teaching the model in code. This insight reinforced Nadella’s belief in the revolutionary power of AI.
Nadella described a personal breakthrough made possible by the GPT 3 model. As a child in Hyderabad, India, he had a lifelong ambition to learn Persian poetry. He was able to fulfil a childhood wish by translating Rumi’s writings into Urdu and then into English using the methodology.
Nadella went on to highlight Microsoft’s long-standing commitment to AI, including its own Turing project, which has been integrated into a number of businesses such as Bing and Azure. Recognizing the same goals, Microsoft has partnered with OpenAI, removing the need for parallel model training and allowing for effective consolidation of resources.
When asked if Microsoft might acquire OpenAI, Nadella was quick to point out that OpenAI is a nonprofit company. Instead, both companies have developed a strategy to create a for-profit entry point that will allow them to reinvest their revenues.
In terms of OpenAI’s pursuit of AGI superintelligence, Nadella emphasized the need to make AI capabilities available to everyone, rather than focusing primarily on a particular benchmark.
The launch of Bing’s innovative chatbot prompted other industry players to incorporate AI into their own products, often prematurely. Some journalists experienced hallucinations or inaccurate responses from the chatbot. Microsoft praised the continued development of the technology, highlighting the creation of novel solutions to hallucinatory problems. The chatbot also offers many settings for users to modify its responses.
When asked about the potential negative impact of AI on the human condition, Nadella acknowledged the dangers of unregulated technology. However, he drew parallels with the processes put in place to oversee power grids and nuclear energy, suggesting that similar efforts could be made to govern AI responsibly.
Levy concluded the interview by asking Nadella if he felt he was being recognized for his achievements at Microsoft, particularly in the area of AI transformation. Nadella respectfully left that decision to others, emphasizing Microsoft’s ongoing achievements and importance in the technology market, with AI transformation playing a key role.