In a recent interview, OpenAI CTO Mira Murati said that the models the company uses in its labs are not far behind those currently available to the public.
Speaking on the release of GPT-4o, she said that it was a massive deal for OpenAI to be able to make this kind of technology accessible to the public. “I don’t think there is enough emphasis on how unique that is for the stage where the technology is today.
“In the sense that inside the labs, we have these capable models and they’re not that far ahead of what the public has free access to. That’s a completely different trajectory for bringing technology into the world than what we’ve seen historically,” she said.
This brings up an interesting point of how, while many AI startups struggle to survive past their early stages, it is still possible to rise up, much like OpenAI did in the last few years.
While Murati’s point is specifically about OpenAI’s ability to make advanced models widely available, she also emphasised that this is to ensure that the general public is fully aware of how the technology is progressing.
“It’s a great opportunity because it brings people along; it gives them an intuitive sense of the capabilities and risks. The opportunities are huge now,” she said. Murati’s stance implies that startups are not far behind in terms of the technology needed to develop advanced AI systems.
Behind OpenAI’s Success
OpenAI is widely regarded as an AI success story. Initially founded as a non-profit towards the goal of developing beneficial AGI, the company quickly rose through the ranks of the startup world to become a household name.
However, there were several factors that helped OpenAI gain the attention it did. Namely, access to funding as well as a large talent pool, allowed the startup to focus largely on quality research.
One of the key points to remember about OpenAI is that its big break with ChatGPT came only a few years after the company received a $1 billion investment from Microsoft. Additionally, the timing was near perfect, as they had focused their attention on the development of Transformer models, shortly after the release of the 2017 paper, ‘Attention is All You Need’.
Of course, it’s key to note that all of these factors rely on ensuring that your problem statement is one that can result in something ground-breaking.
What Needs to Be Done?
Murati’s implication that AI models are much more widely available now than ever before, means that startups need to try harder to stand out. OpenAI appeared as a pioneer in making an easy-to-use AI product widely available to the public for free. However, only years later, doing the same doesn’t have the same impact anymore.
In India, several startups have managed to make a splash, albeit not as big as OpenAI’s. Successful Indian startups like Krutrim, Kissan and Sarvam have been able to leverage gaps between the AI revolution and use cases specific to India, allowing them to break ground as pioneers in the country.
Being able to identify certain gaps and plug them in a way that ensures a startup’s continuity is a tough task. This is especially true as rapid advancements from bigger companies mean that startups with a poorly thought-out plan and product find themselves obsolete almost overnight.
With the proper resources, however, this can be remedied. One must keep in mind that OpenAI’s ability to get ahead of the curve was solely due to a team of researchers that tried to leverage then-unexplored technology for use among the general public.
AI advancements mean that a lot more students and professionals are pivoting towards studying AI, thereby creating a large talent pool, especially in India. Besides, with startup funding increasing exponentially, it seems that the only roadblock now is trying to find a way to stand out in a sea of problem statements.
However, as Murati said, “If you can push human knowledge forward, you can push society and civilisation forward.”
