What OpenAI Means for the Entire Tech Sector
When ChatGPT launched, exactly a year ago, it demonstrated the potential of the first new general purpose technology in a decade. It was the most important development since the unveiling of the iPhone, which triggered the ubiquitous use of the internet away from a desk.
With OpenAI’s rise, the first new tech platform in the West, since Facebook, has emerged and its entry into the landscape populated by Google, Apple, Meta and Amazon will affect the relative standings of them all. Amazon will be affected least directly and Meta will be protected as its content is largely behind a walled garden. Meta will also act as a hedge against OpenAI’s dominance as it is set to gain the most if open source LLMs win against proprietary ones. Microsoft’s 49% stake in OpenAI is self explanatory in terms of its benefit or threat and the impact on Apple may be largely symbolic but nevertheless important as it ceases to be the brand that people get excited about. For Google, the threat is more direct. Here we’ll look at how.
Yesterday was an equally important event to the ChatGPT launch, in several ways. Firstly it marked OpenAI keynotes taking the mantle from Apple. In terms of buzz and excitement, the announcement of a new charging cable for a nearly identical phone doesn’t cut it compared to being able to build an app on top of all the world’s publicly accessible knowledge, via chat.
Apple has squeezed the juice of iPhone revenue and delivered fantastic results, but it has failed to innovate. In the decade since Jobs died it has delivered a pair of peripheral devices to the iPhone — a watch and earphones. The Vision pro feels less like a product, form factor and price that people want vs wishful thinking by Apple of what a device to replace the iPhone could be. Apple has admittedly innovated under the hood with proprietary chips, but this leads to selling on numbers like more memory and faster processor speed, something that was historically associated with PCs rather than Apple’s qualitative design excellence. Such selling on numbers was something Jobs explicitly avoided where possible. On the outside, Apple hardware from its product cases to retail design hasn’t changed in years. For standalone software outside of the OS, none of its applications, from mail, to maps to productivity tools are best of breed. Increasingly, its 30% rake on fees and in-app purchases seems greedy and could be challenged and its features like live photos, which use three times the storage of regular images, look cynical.
Companies like Apple and Microsoft go through cycles of innovation and recouping the rewards of that investment. For the last decade, Apple has been in a mode of efficiently harvesting the benefits of a previous period of innovation, rather like Ballmer’s period at the helm of Microsoft. But eventually Microsoft needed another innovator and that came in the shape of Nadella who achieved a new phase of innovation through incredibly smart acquisitions — including half of OpenAI. Apple feels like a company that needs a new innovator, particularly now that the landscape has been shaken up with generative AI and its own forays into the space, such as with Siri, are widely perceived to be inferior.
In summary the buzz around Apple feels like it has migrated to OpenAI and the GPT app store potentially extends the app store model to everyone.
With the Announcement of the GPT marketplace OpenAI now have a revenue model that scales beyond the freemium SaaS one, by leveraging third party developers. By making anyone a developer and then incentivizing them through revenue share, they scale that to an unprecedented level.
Although similar in product terms to the launching of the iPhone app store subsequent to the iPhone launch itself, in business terms, this is similar to the trajectory of Google which found a revenue model by copying Overture’s paid placement in search results, having launched an innovative product without a revenue model, first. And it is Google’s business model that GPT applications challenge head on.
Google’s business model was already challenged by generative AI, since it relies on user generated content as raw material, from indexing the web to streaming user submitted content on YouTube. In many ways the web in 2022, which formed the basis of ChatGPT represents a data set that is analogous to the 11th edition of the Encyclopedia Britannica, which was noted for its primary sources with entries written by people such as Bertrand Russell or Ernest Rutherford. After 2022 an increasing portion of the web will not consist of sources that are primarily human and so it potential quality both for search and training may reduce.
The introduction of a business model where people can monetize any proprietary dataset effectively extends this further and puts up a paywall across all data. It has potentially created an incentive to not publish useful content on the web.
With its new focus, OpenAI becomes like a black hole that will suck anything monetizable across a data event horizon, eating into Google’s raw material and turning the web inside out.
A new infoverse has been created!
