The Cost of Distribution Goes to Zero
Here’s a problem very few people have had in recent years: newspaper ink stains. Pouring through the Sunday Times inevitably meant having to wash a thin black film off of your fingertips… but when was the last time you held a physical newspaper in your hand long enough for it to stain your fingers?
The internet had a profound impact on the competitive dynamics of a variety of industries (publishing included—making newspaper ink stains a thing of the past for a large percentage of the population). One it’s key impacts, captured in Ben Thompson’s Aggregation Theory, has been to lower the cost of distribution for digital goods and services to zero:
Aggregation Theory suggests that companies that control the most important distribution channels in an industry will have a significant advantage over their competitors. In the internet age, this often means owning a platform that aggregates content or services from multiple sources.
For example, in the retail industry, Amazon’s platform aggregates products from millions of third-party sellers and provides a seamless shopping experience for customers. This has enabled Amazon to become the dominant player in e-commerce and create a powerful network effect that makes it difficult for competitors to catch up.
Similarly, in the ride-sharing industry, Uber’s platform aggregates drivers and riders and provides a convenient way for people to get around. This has allowed Uber to disrupt the traditional taxi industry and establish a dominant position in many markets.
The internet meant the cost of distribution was next to nothing and value accrued to the platforms that could most successfully ‘aggregate’ attention and customer relationships. This has been the competitive reality for internet-based businesses for the past decade. Yet, although distribution costs have more or less been eliminated, costs of production were often still significant.
The Cost of Production Goes to Zero
- The New York Times compensates hundreds of journalists tapping into thousands of sources to create their content.
- Fashion houses spend millions per year on sets, designers and models to get their brand imaging and product positioning just right.
- Netflix spent around $16 billion on the production of content in 2022.
This might all change thanks to the rapid progress of Generative AI—and specifically, large language models (LLMs) like GPT— and their deployment of across all corners of the web. Josh Brown illustrated this perfectly in a post over the weekend:
How many people don’t know how to code? Most?
We’re heading into a future in which not knowing how to do a thing is going to matter much less than ever before. Technical skills will be less important than creative skills. How will matter less than why. The ramifications for entrepreneurs in all industries, including mine, are unfathomable. The ground is quaking.
What would you create? What would you build? If you could will an idea into existence simply by typing a natural-language command into a chat box and having trillions of dollars worth of compute power make it so – at no cost or expense of time to you – what would that feel like?
The swift spread of GPT and other forms of Generative AI have meant the cost of production for digital goods and services is also inevitably going to zero. Anyone anywhere now has the power to spin-up products and services assuming they are based in an element of code, text, image, video and/or audio. This will lead to an explosion of content creation, small apps and businesses as noted in the latest Not Boring essay:
ChatGPT both makes it easier for people to build Small Apps – it can create a barebones website from a sketch, or walk users through how to code up sites – and makes it easier for those Small Apps to find customers as long as they differentiate hard enough on certain attributes.
This shakes the foundations of aggregation theory. The act of ‘Creation’ has shifted from scarce to abundant.
In a world of unlimited distribution, unlimited production means it will be harder to find signal through an increased volume of noise.
So where is the new scarcity?
Brand is Where Value Will Accrue
Brand becomes what is scarce. Trust, reputation and accrued experience is the new value chain component that will aggregate users.
- It costs a company nothing to mint 200,000 #NFTs. It costs a company everything to create the #brand and positioning required to create demand for those NFTs.
- It costs Reid Hoffman very little time and energy (at least, less than a typical author) to create his 240 page book about AI. It costs Reid Hoffman everything to build his personal brand into what it is today.
- It costs Morgan Stanley nothing to produce GPT-driven financial advice. It costs Morgan Stanley everything to develop the perceived staying power and competence for clients to confidently act on those recommendations.
Similar to how platforms became dominant in the Aggregation Theory age through the accrual of attention, brand will become dominant in the coming Generative AI age through similar means. In fact, brand was already an accrual mechanism. Brands act to aggregate everything positive and negative a company has ever done into one all encompassing signal that consumers use in their decision-making processes.
This becomes obvious through a contra-example. Picture for a second a world without big brands where goods are no longer mass produced on a one-to-many basis, but rather produced by craftsmen on a one-to-one basis. How would you evaluate that craftsman’s quality? How could you compare price points between craftsmen? How would you know they would honor a return or guarantee? How would you know they wouldn’t walk away with your money today and never return?
These are all things that accrue to a brand today. They act as a ‘check’ on companies to encourage them to do things that are in their best reputational interests. They are also an all encompassing information package that consumers can use to decide between providers in a market.
Who Gets the Brand Advantage?
In a world awash in Generative AI, brand as a moat becomes stronger than ever. Rex Woodbury argues this will cause big brands to become more powerful than ever at the expense of small brands that will have a tougher time fighting for consumer attention.
In his Not Boring essay, Packy McCormick suggests there is some nuance to this dynamic, particularly as it relates to the introduction of GPT Plug-ins, which was announced a couple of weeks ago as a simple plug-in for developers to be able to embed GPT in third-party services. In this case, separation won’t necessarily be created between big brands and small brands, but rather between differentiated brands and undifferentiated brands.
The takeaway as I see it is that, again if OpenAI chooses to go this route, a lot of industries and value chains are going to be shaken up, and the winners will be the companies that focus on differentiation, on doing a specific thing really, really well, as opposed to those who do a lot of things pretty well and pump money into brand. Of course, this won’t be true in every category. Luxury fashion will still be valuable as a status signaling mechanism. Software products that organize complex workflows for companies will still retain a lot of customers and may find a lot of new ones through ChatGPT.
I’d argue there is a third way of looking at this: rather than big vs small, or differentiated vs undifferentiated, we will also see a separation between fast brands who are able to keep up with shifting consumer preferences and the changing advertising/distribution/creation landscape and slow brands who are not able to respond fast enough.
We have already caught a glimpse of brands that are experimenting with GPT Plug-ins. Familiar companies like Expedia, Instacart, KAYAK, OpenTable, Shopify, Slack, and Zapier are among its first users. We can also see the personal brands and influencers who are experimenting with GPT (ie. see the Reid Hoffman example above or this example of David Sacks creating a blog post).
GPT and Generative AI will amplify the speed advantage present in most modern-day industry dynamics.
- Those that can adapt to the new landscape fastest will be in the best position to break-through the noise that will inevitably be created in a world of limitless content creation possibilities.
- Those that can position themselves within GPT correctly will rise to the top of the conversational pile of AI-generated content and recommendations (similar to how Google and search spawned an entire industry around search engine optimization (SEO), we can be sure that with 100+ million users already, GPT Optimization (GPTO) will be a similar opportunity).
- Those that can adapt their content and messaging to align with shifting consumer preferences the fastest will have the greatest chance at remaining on-trend, top-of-mind and culturally relevant.
In each case, the value of being ‘fastest’ accrues to the brand itself and amplifies it as a competitive advantage.
One AI to Rule Them All?
There is an obvious sidebar to this conversation: Value will also clearly accrue to LLMs and the companies that produce them. Much has been written about Generative AI as the next ‘technology platform’ and a few months into the launch of GPT and the release of Plug-ins, that appears to be the path we’re headed down.
But what happens if one AI rules them all? What happens when one model drives everything. Do we revert back to a world of sameness. Tadas Viskanta recently raised this as an issue around the ‘index mindset’ and how in varying domains ranging from coffee shops to HGTV-driven home design, everything now looks pretty much the same:
The index mindset is comfortable – avoiding decisions requires the least amount of effort. But if you index across every domain, you lose your differentiating features, becoming an average of everyone else. Online tools undoubtedly make this issue more pervasive. The widespread introduction of a new class of AI tools will only make it worse. The challenge in broad, real life is we don’t necessarily know when indexing works and when it doesn’t.
Brands that are able to adapt and know when/how to ‘stand out’ versus ‘fit in’ will also have advantages in the new world this AI platform shift presents.
Yet, there is reason to believe that there will be more than one platform to build on. Today’s big tech firms are not going to let OpenAI runaway with the lead. Bard is making progress over at Google moving from LaMDA to their larger PaLM model. Baidu is pushing forward with their Ernie Bot. Meta has opened up about its LLaMA model. Bloomberg released BloombergGPT, a 50-billion parameter large language model, purpose-built for finance. Stability AI continues to make forward progress on text-to-image and generative animation. Other AI-based start-ups have run with text-to-speech and text-to-video. The pace of progress is astonishing.
Much like Apple and Android, developers will have options—but likely not many. Platform business models tend to have winner-take-all effects and the AI shift will likely be no different. However, it is too early to tell who those winners might be.
Conclusion
In a world of Generative AI, the noise to signal ratio will become unbearably difficult to navigate. Brand becomes our new navigation mechanism.
It is shocking to me how fast things are moving. If I haven’t shouted them out already, The Neuron Newsletter put together by Pete Huang and Noah Edelman is a fantastic resource for staying on top of the steady flow of news, company launches and perspectives in the space.
Brand has always played a role in how consumers make decisions. With the cost of distribution dwindling to near-zero thanks to the internet, and the cost of production now falling thanks to Generative AI, it is clear that brand will play an even more prominent role in consumer decision-making going forward.
Those that are able to find advantage through size, differentiation, speed, or their ability to ‘read the room’ by standing out or fitting in will emerge as tomorrow’s leaders.
Paper & Blocks 