AI Will Make Big Techs Even Bigger
tl;dr
Generative AI won’t significantly change the leadership position Big Techs occupy today.
Amazon, Google, Microsoft, and others were quick to adopt different Gen AI strategies to capitalize on this new technology.
Investments in leading Gen AI startups, developing models internally, and adding Gen AI-powered functionalities into their existing products are some of the measures taken by these players.
Regardless of their strategy, Gen AI is heavily dependent on hardware and cloud services that Big Techs currently monopolize.
This month, news broke out that Google is reportedly investing up to $2 billion in the Generative AI startup Anthropic. This was just weeks after Amazon announced they were investing $4 billion in the same startup, making it clear that Big Techs (Google, Apple, Facebook, Amazon, and Microsoft) are rushing to capitalize as much as possible on the AI hype and ultimately not fall behind competitors. Each company has a slightly different strategy to achieve this but I believe their competency, added with a bit of luck, will be enough to maintain Big Tech’s dominance in the long term.
Different Versions of the Same Story
Before giving my arguments as to why I don’t think Big Techs will lose substantial relevance in the “AI revolution”, let’s look at each player in this group and analyze their Gen AI strategy.
Microsoft
Investment + Product Functionalities
Microsoft was the pioneer among Big Techs to make moves in the Generative AI space. The company first invested in ChatGPT’s creator OpenAI back in 2019 with a $1 billion check. In the following years, Microsoft doubled down on that bet, investing additional billions into the startup up until the official launch of ChatGPT in November 2022. Soon after the meteoric rise of the tool, Microsoft reportedly agreed to invest up to $10 billion in OpenAI and announced it would implement AI into almost all its productivity software products, including Word, Excel, PowerPoint, and Outlook. As part of the investment, Azure, Microsoft’s cloud platform, became OpenAI’s exclusive cloud provider, powering the startup’s workloads across research, products, and API services.
Google
Investment + Product Functionalities + Develop Internally
Google is taking a slightly different approach when it comes to its Generative AI strategy as compared to Microsoft. The search giant has been conducting breakthrough AI research for decades and they rushed to launch Bard, their conversational AI search tool, as a response to ChatGPT. Bard is powered by PaLM 2, Google’s own Large Language Model (“LLM”), which is the name given to the foundational models that power Generative AI applications such as ChatGPT. The same model is also powering all AI functionalities that they added to all of its workspace products, including Docs, Sheets, Gmail, and others. Internal efforts haven’t stopped them from also making external bets in the form of their $300 million investment in Anthropic in late 2022, followed by their $2 billion investment in the same company.
Amazon
Investment + Product Functionalities + Develop Internally
Amazon is adopting a very similar strategy to Google when it comes to Gen AI. They recently announced that they will supercharge all their Alexa devices with their own LLM, offer generative image capabilities to their advertisers, and use Gen AI to make their internal processes more efficient. The company also invested $4 billion in Anthropic, the same AI company Google invested in, and, as part of the investment, the startup agreed to use the e-commerce giant’s chips to train and deploy their model as well as use AWS as their primary cloud provider.
Meta
Develop Internally + Open Source
Meta (formerly Facebook) is adopting a different Generative AI strategy compared to its peers. Instead of developing its internal models, keeping the source code private, and charging whoever wants to use them, Meta is following an open-source model. This means that everything they develop is open to developers to work on top of and commercialize however they wish. This approach was praised by the developer community and goes against the closed-door approach from most of Meta’s peers.
Apple
Develop Internally + ???
Differently from all its peers, Apple has been quiet about its Generative AI strategy. Tim Cook, the company’s CEO, has mentioned that Apple implements AI and machine learning-powered features into all its products but they made no public announcement specifically about their Gen AI efforts. There have been some reports that there is an internal team working on Apple’s LLM but it is unclear how they will implement it into their existing and new products. Interestingly, according to CB Insights’ data, Apple is leading the way in terms of AI-related acquisitions which indicates they might not be as behind as it may seem when it comes to AI adoption.
Why Big Techs Will Still Win
It is difficult to predict which of the above strategies will be most successful in the long term but one thing is clear for me - Generative AI won’t significantly change Big Tech’s leadership position.
The first, more obvious, and weaker reason for this is because of their heavy investment in the leading players in the Gen AI market. If those foundational model players such as OpenAI and Anthropic, do indeed end up becoming long-term successful AI companies, Google, Amazon, and Microsoft will probably try to acquire them entirely or in the worst case see huge returns for their already huge investments (which won’t significantly move the needle for any of them).
The second, and most important reason, is that Generative AI companies are heavily reliant on Big Techs. Foundational models that run ChatGPT or any other Gen AI application require huge amounts of computing resources (GPUs and CPUs) to be trained and Big Techs control most of the computing power available. GPT-4, for instance, was said to have cost over $100 million to be trained with a big chunk of this money going to Microsoft, Google, and Amazon, who collectively control around 65% of the world’s cloud infrastructure. As a proof point, Microsoft’s most recent earnings release highlighted a higher-than-expected growth of its cloud business.
The final reason, which is still to be defined as how strong it will be, is that Big Techs were quick to implement Gen AI into their existing product base. Any startup trying to launch a Gen AI-powered workspace tool, for instance, will probably be quickly outpaced by those companies’ huge distribution capability. Brand new AI-powered products could also turn out to be big contributors to Big Tech’s longevity. Less than 1 year after its launch, GitHub Copilot, a product owned by Microsoft, reportedly crossed the 1 million user mark, which, multiplied by its $10 per month pricing, means it generates $120 million per year for the Big Tech. This is still not a significant value for the company but still shows impressive growth.
Every major technology revolution has historically produced new leading companies that have entirely or almost entirely replaced incumbents. The AI revolution will probably continue this pattern with one key difference. Newcomers will most likely not entirely replace the more traditional companies but actually share leadership with them. Without a democratization of computing access and commoditization of foundational models, it is simply hard to imagine a future in which Big Techs aren’t relevant.
Dive Deeper
The Future of AI Is GOMA by The Atlantic
We cannot allow AI to make Big Tech even bigger by Steve Case
Connectivity, Global Fragility and the Added Danger of AI by Nassim Nicholas Taleb (video)
Thought of the Month
I’ve been re-reading Peter Thiel’s “Zero to One” and I wanted to share his view about the importance of “secrets”. It might be helpful to anyone trying to find a good idea for starting a business.
“…what valuable company is nobody building? Every correct answer is necessarily a secret: something important and unknown, something hard to do but doable. If there are many secrets left in the world, there are probably many world-changing companies yet to be started.” - Peter Thiel.
Chart of the Month
According to a study from Capital Economics, rich nations are positioned to benefit more from AI compared to developing economies. It looks like AI will not only make Big Techs bigger but also rich nations richer… The big takeaway is we need to work hard to make sure AI goes in favor of equality and not the other way around.