What’s our window on our tech-centric world? We know what it is today; the smartphone. Will that be true in three, five, ten years? Elon Musk, for what it’s worth, says that smartphones will be gone in 5 or 6 years, replaced (he thinks) by direct AI links to the brain or advanced wearables. Enterprises who have a view on that question divide into some fairly distinct camps, and why that is and what they are is interesting and important in our personal assessment of the future.
To get a point out of the way, I can’t say that it would be impossible to set up an AI-to-brain link at some point in the future. Today? No. In 5 or 6 years? Only in some very basic way, I think. Enterprise enthusiasts generally agree with that, and the pragmatists think it’s pure science fiction. Some point out that early sifi novels predicted intelligent robots by the year 2000 (I remember one such novel, “The Door into Summer” by Heinlein). In any event, wild predictions get you publicity but are rarely useful for real tech planning. Let’s try to get a bit more realistic.
If we presume, as I do, that the biggest apple remaining for tech to pick is real-time, real-world empowerment of workers and consumers, then it follows that tech would have to gain a very good insight into the conditions of the real world around each user, and in real time. That would mean having a world model with a decent level of detail, because the extent to which tech could “see” the world around each of its users would determine the value it could add to work and life.
There would, I think, have to be two levels of model, one a general model that everyone in a given “locale” used, and was populated by generalized sensor/camera technology, and another private model that each user populated and was theirs to control access to. The most logical source of data for the private model would be a camera embedded in XR glasses, because AI analysis of video can provide more information faster than any other mechanism.
The other side of the picture is that our AI would need to be able to influence the behavior of the user. OK, if you’re an enthusiast with a very long timeline, you can assume that it could exercise mind control and thus movement control, just as you can assume that the user could use brainwaves to control AI. If you want a bit more pragmatism and a shorter path to realization, you have to assume that the influence would be expressed via XR, simply because (as I’ve just noted) there’s no information path to the brain as capable as vision.
Enterprises see two different pathways to achieve this sort of thing, both driven by business. One path focuses on employee empowerment, and the other on consumer empowerment.
The employee empowerment push is aimed largely at reaching the roughly 40% of the workforce who aren’t working at desks or fixed positions like cashier lines, toll booths, etc. This group has a slightly higher total unit value of labor than the 60% we already manage to support, but the challenge is that their jobs are highly varied, and just what they’d need is highly varied as well. That makes it hard to put a framework into place that intercepts enough value to create an ROI.
Consumer empowerment is also technically complex, but it also has a problem of overall mission. It’s easy to say it, but hard to identify the specific path to monetization. Remember, to get companies or people to invest, you have to define value that they’re willing to pay to obtain. Consumers are conditioned by the Internet to prefer ad-sponsored, to-them-free, offerings. They are resistant to actually making direct payments for anything they perceive to be a part of the online ecosystem. The problem is that ad sponsorship is that it’s a small fraction of the global market for goods and services—my model says it’s $1.2 trillion in a world GDP of $116 trillion. We clearly have ads these days, and the fact that most streaming services are offering direct-pay options for ad-free streaming shows both that some consumers are weary of ads and would resist more interruptions to their lives, and that some sellers see ad revenues declining as the number of places where ads can be placed proliferate and competition for eyeballs increases, driving down prices.
So what do you “empower”, exactly? Many think you need to focus on the shopping process, perhaps actually involving tech in sales. A person walks past a product display, is notified that it contains something they’re shopping for, and offers an opportunity to buy it in passing by, without even entering the shop. Or, of course, the shopper might simply be told where the cheapest price could be obtained nearby, or along a route regularly traveled, when the desire to buy was registered. You can see that a retail connection isn’t automatic, but Amazon already embeds “ad to cart” ad click options to viewers of Prime Video, so it works at that level. That suggests that some form of consumer/XR purchase empowerment would also work.
Consumer empowerment, like worker empowerment, is likely to depend technically on world models, but unlike work, which tends to happen mostly in concentrated locales, consumers roam about everywhere and thus their empowerment might logically expect to need more comprehensive world models. Combine this with the risks of even deciding how to target empowerment, and it’s hard to see how consumers could lead the charge in tech advance, because of the high entry costs to providers.
All this is important because what we advance to in terms of tech’s way of interacting with us (and vice versa) depends on what we want it to do, which depends on what some set of sources is willing to invest in. I think that it would be possible, even perhaps easy, for some business empowerment to proceed on a per-company basis. I think that the consumer side might do the same if it focused on retail mall environments first, but the problem I see is that what consumers want is the lowest prices, which is not what sellers want, and sellers have to buy into any retail targeting of technology because they set the prices and implement the order processing, warehousing, etc.
The challenge for both, and for tech overall, is that increasingly obvious challenge of ecosystem-building. XR glasses, world models, low-latency networking, edge computing are all likely necessary conditions for both these real-world missions, but none are sufficient conditions. That creates an inversion of the popular first-mover advantage paradigm; the first mover here is on their own until everyone moves.
But to get back to Elon and smartphones, I think he’s totally wrong. Even with smart devices like watches, rings, and glasses, you need a lot of not-yet-available technology to get these to connect with the real world, unless you presume the phone is there as an intermediary. I think that rather than eliminating phones, the real-world evolution in the worker and consumer spaces will empower them.
