Microsoft, Google in AI Proxy War
What’s intriguing about this large scale largesse is that it is limited to a handful of startups
In a technology age where those who build are often bought over by those with fat purses, having just two words on an elevator pitch seems to be enough to become attractive. The two words are artificial intelligence (AI) and existing big tech companies have pumped in a whopping $12 billion (with more committed) into just two companies – OpenAI and Anthropic.
Google recently invested $2 billion into Anthropic, with Amazon committing to $4 billion more to narrow the funding gap that saw Microsoft throw $10 billion into OpenAI, the company that created a “shock and awe” moment last November by coming out with ChatGPT, a chatbot that could potentially take away the creativity from human existence.
Of course, the deal between Google and Anthropic, reported by the Wall Street Journal, indicates that just a quarter of the commitment would be met now and terms and conditions, if any, are not known. Be that as it may, the fact remains that it’s pay-day for companies that can lay claim to AI leadership.
AI isn’t new and others can use LLM too
This does not mean that there aren’t others with the AI tag doing solid work in this space, which is not as big an innovation as some would have us believe. Recently, a former colleague took pains to explain that a self-switching turn indicator on a four-wheeler was also artificial intelligence at work. If the industry is so smart, how come this tech isn’t available on bikes?
Given that both OpenAI and Anthropic work around large language models, the probability of another startup announcing an incremental feature to the existing set remains high. However, the race to invest in anything AI among the Big Tech merely suggests a burning desire to be part of, or actually own) existing AI-led platforms. Not being a part of this innovation is something that could queer their narrative in the future.
What are these AI companies gaining?
From the point of view of OpenAI and Anthropic, hard cash is a prerequisite for setting up the required infrastructure (powerful data centers) that would be required to deploy new AI models and test them at scale. Without the funds, these companies would be scratching around with little to no revenue opportunities to monetize their innovation.
Now, if one were to ask why these companies aren’t keen to invest in a single entity, the answer could be a two-pronged one. For starters, they could end up becoming customers of AI-led smarts held by another entity. Also, there’s the big question of bragging rights whereby new innovations could substantially enhance their own values among stakeholders.
Are all AI companies solving the same problem?
So, does this mean that these AI startups (to call OpenAI one might be an injustice) are all working to solve the same problem? Probably not. Because Anthropic seems to focus on AI models that are less interesting to you and me but more so to enterprises seeking long-term business models for growth.
While OpenAI is churning out new features that includes creating voice and images using their AI models, Anthropic claims that its sweet spot revolves around knowledge work and professional services. We may need to wait some more weeks or months to figure out what its CEO and co-founder Dario Amodei has in mind.
The company also claims that it has data security, intellectual property and transparency high on its list while creating AI models. So, writing a movie script or an essay on sustainability isn’t something that it wants to get into (OpenAI does so via ChatGPT). Instead it wants to create solutions that corporate customers recognize as game changers before signing up.
A proxy war to establish the right narrative
Finally, the question that this all-out proxy war throws up is where are the funds going? For a long time, we’ve known that aI models are expensive to train and by the time a company squeezes one to a point where it costs a dollar to answer a query, a more powerful model may arrive. OpenAI may be free, but Anthropic is clear that it would offer nothing for free.
How these models work and what the Big Tech companies would spend on them or possibly develop internally is for time to tell. These are questions that even OpenAI and Anthropic may not be able to answer using their smart chatbots.
