In the past few weeks, Google has rolled out scores of AI updates. It enabled AI-generated podcasts, the latest DataGemma model to address AI hallucination, improvements to Gemini 1.5, and more AI features for content creators. Logan Kilpatrick’s cryptic tweet about tokens has also sparked speculations about Gemini 2.0 coming soon—possibly by the end of the year.
Google is aggressively shipping more products, that too in a short span.
The Silicon Valley powerhouse has always been ahead of the rest, especially OpenAI, in terms of research, tools, infrastructure, and resources, to materialise AGI. However, what it lacked was an entrepreneurial ecosystem where this vision could have been advanced.
For context, In 2017, Google’s researchers released a paper titled ‘Attention Is All You Need’, that changed the world of modern generative AI as we know it. The ‘T’ from the GPT, aka the Transformer model—was the breakthrough architecture—later employed in OpenAI’s ChatGPT model since its inception.
Many of the researchers from this team either went on to build their own ventures, or later joined OpenAI’s team.
Google is the New OpenAI
Since the recent launch of o1 and OpenAI’s recent $150 billion valuation speculation, the company has far surpassed its startup status. Chief Sam Altman’s dramatic ousting and return in November 2023 was also the start of the organisation’s shift in priorities.
Back in 2015, Elon Musk, Greg Brockman, and Altman founded OpenAI as a nonprofit AI lab with a singular goal of AI for humanity, without any profit maximising incentives. However, Altman’s vision for OpenAI kept shifting towards a more traditional for-profit model, as the pursuit of AGI needs funding that would not be feasible under its original charter.
OpenAI has definitely come a long way from where it began. Recently, Kilpatrick told AIM about how his journey from OpenAI to Google felt like going back to a startup. “When I joined OpenAI, it was a 100-person company, and when I left, it was much larger — about 1,500. In many ways, coming to Google has felt like going back to the roots of that 200 person startup,” said Kilpatrick, stating how Google—despite being big—feels like a startup, housing multiple startups within the organisation.
Meanwhile, OpenAI, on the other hand, seems to be taking a more decentralised approach, where there is a clear bifurcation between the product and research teams, both of which work independent of each other, hyper-focussed on their specific tasks. Google’s culture is more collaborative, where the DeepMind team actively engages with developers and researchers.
“Gemini is related to twins, and I feel twins is a good name for the legacy of both—Google DeepMind and Brain—to come together and work on ambitious, multimodal models,” said Jeff Dean, chief scientist of Google DeepMind and Google Research, on the inception of Gemini, that was launched in 2023 as a competitor to OpenAI’s fast-growing ChatGPT.
Google is Now Shipping Faster Than OpenAI
With the launch of o1, OpenAI has entered level two of the five levels of AI’s revolution. This post on X is indicative of OpenAI’s departure from its earlier fast-launch or ‘ship fast’ approach towards a more well designed, focussed, and functional approach that would help in the long run, and ultimately in achieving AGI.
“With LLMs in particular, we invented them basically. The Transformers research paper for instance we were perhaps too timid to employ them. They could make embarrassing mistakes. At the same time, AI can do powerful things. The capability of AI is magical. I think of it as something magical we are giving to the world. I don’t think this is a technology you should keep close to your chest, and hidden, until it is perfect,” said Sergey Brin, former president of Alphabet, in a recent AI Summit.
Google, on the other hand, is moving further away from the approach of obsessing over the ‘perfect’ model. This is similar to how OpenAI functioned in its early days, where the focus was to launch as quickly as possible.
“To me, we are not going to end up in a winner-take-all situation. I think the players in the market understand how much of an opportunity there is with this technology and how transformative it is. If anything, it has pushed more people to compete in this space, which is a net benefit for everyone,” said Logan in an interview with Cognitive Revolution podcast.
Google’s I/O event earlier this year was also a reflection of this trend. A lot of people believed that it was a series of fragmented and minor updates, instead of the one big update.
