Yesterday, OpenAI unveiled its latest model, marking a significant leap forward in AI technology. Today, Google is expected to announce Gemini updates at their I/O keynote. Here are some insights on what these developments mean, where the technology is headed, and the implications for business leaders in an evolving landscape.

First, what happened?

Dubbed GPT-4o (the 'o' stands for omniscient), this model introduces groundbreaking enhancements in speed and multi-modality. Those familiar with ChatGPT's voice capabilities will appreciate the improvements. Personally, I use it to allow my 5-year-old son to explore his curiosity. He asks questions on diverse topics and delves as deep as he likes, receiving answers in a natural, human-like voice. Previously, the response time was 3-5 seconds, but GPT-4o cuts this down to an average of just 0.3 seconds—mirroring human interaction times.

Moreover, GPT-4o can now simultaneously process and reason across audio, visual, and textual inputs in real time. This positions GPT-4o as an interactive co-worker in knowledge-intensive tasks. We strongly recommend checking out a few demos, for example on math tutoring and coding, to grasp the extent of this advancement. If you can, OpenAI’s 26 minute presentation is well worth watching. OpenAI has also launched a desktop app for Mac, with a Windows version on the way, paving the path to a future where a highly intelligent aide is ever-present. 

Second, where is it going?

The trajectory of AI development is clear. Each new model significantly outperforms its predecessor, with groundbreaking improvements expected every 4 to 8 months, as noted by Dario Amodei, former VP of Research at OpenAI and now CEO of Anthropic. Future models will reduce errors (hallucinations), enhance reliability for specific company data queries, and introduce agents capable of executing complex, multi-step tasks.

However, the financial outlay required to develop these models means only the most well-funded entities can compete, creating a feedback loop where success begets success. The consolidation trend is already apparent. For instance, the recent folding of Inflection AI into a quasi-acquisition by Microsoft. This is consistent with past technological waves - for example,  Google in search, Apple in smartphones. Rumors of an OpenAI partnership to integrate ChatGPT with iPhones (imagine Siri, but vastly improved) suggest consolidation is accelerating.

Implications for Business Leaders

The advent of GPT-4o presents several implications for business leaders. Firstly, leaders should engage directly with these tools. The use of generative AI has nearly doubled in the last six months, with 75% of global knowledge workers now using it, according to Microsoft. Leaders should explore these technologies firsthand to uncover advantages for their organizations. Secondly, with the capabilities of GPT-4o, and more advancements to come, now is the time to identify and leverage use cases that can benefit your business. Lastly, strategic vendor selections are important; prioritize collaborations with vendors that are likely to endure. When selecting AI model providers or applications infused with AI, the primary criteria should be the quality and relevance to your use case, but the long-term viability of the provider must be considered. We advise particular caution when dealing with smaller vendors that rely on their own proprietary models.