TLDR: Magic = 10X improvement + autonomy.
What does it take to be magical?
Sometimes you work with people (the maniacs) who could be described that way, or who at least have the makings of autonomous autodidacts (the winners).
But the same is true of experiences with products and brands.
When products are 10X better like ‘the maniacs’ or when they’re self-reliant like ‘the winners’ in the links above, you find yourself experiencing something that can only be described as magical.
Not to totally anthropomorphize the product experience, but think about it, when you use something—like the iPhone when it first came out—it was magic.
The ability to have all of your music, access to the internet, text message, call, and download applications was a step-change experience over the alternative phones. The device had not only achieved much more—like the maniacs—but it seemed to operate in almost frictionless way too, making it feel almost self-reliant—like the winners—too. The intuitive gestures of the touch screen, swiping and pinching for example, were so innate that even a child could pick up the device and use it to some degree.
Those things all became possible though, not as inventions thanks to the iPhone itself, but rather the iPhone became possible as the byproduct or second-order effects of other things changing. Sizes of batteries, quality of glass for touch screens, memory cards, internet bandwidth and access, and so on. All of these things had to improve simultaneously for a product to be perceived as a frictionless step-change.
We’re seeing the same things now in AI.
We have one perceived major breakthrough that was brought on by a series of other breakthroughs and technology improvements. ChatGPT and Large Language Models (LLM’s) are the perceived breakthrough, when in reality it is the decades-long work developing transformers, faster and more powerful compute, few-shot learning, reinforcement learning through human feedback, energy efficiency, graphics solutions and processing products, and so on.
Yet the products being built on top of these systems have yet to deliver true magic.
They have one half of the equation. The magic without the self-reliance, or at least the frictionless experience.
Most all AI products the mass market has been using or raving about in the past six months are simply point solutions that execute a feature. They summarize call notes, autocomplete some text, help draft a PowerPoint, autofill a spreadsheet, or help you interact with your own data more efficiently.
But to build something that is both a step-change and frictionless is a tall order. These products, instead of being consumer point solutions to execute a task need to evolve into interoperable or end-to-end software experiences where a multi-step process is reduced as close to zero as possible.
Either interoperable in that the agents are working with the other agents — think a smart customer data platform for example, where not only are datasets blended and made actionable, but the actions themselves are teed up for one-click execution. This would turn executive leadership for example into a true decision-making function rather than a people management one.
Or if not interoperable, then at least end-to-end, where a verticalized solution builds the entire experience for a set of problems—like a financial advisor who needs a more intelligent customer management platform, or a real estate broker who needs a similar thing, but with very different functionality.
The bar for magic is higher than most founders or consumers realize, and that’s why we have so many products that are just ok, but if that product is already a step-change in functionality, then the next job is to pair it with autonomy.
Like the iPhone succeeding on the heels of second-order opportunities created by a myriad of industries evolving simultaneously, I think we are on the precipice of something similar in energy costs, longevity, and AI.
With nuclear fusion and energy costs approaching zero at the same time as longevity science is maturing to increase healthspan and lifespan, we have a series of second order effects to consider, and when you add AI and the unpredictable timeline of AGI, it’s hard to think of a single industry that isn’t impacted in a material way.
The question is what will emerge that goes beyond simply making something better, but makes it so much better that the experience itself is frictionless and autonomous?
See you Monday.