Back

In the #1980s, personal computing took off with the #IBM #PC, #Commodore64, and #Apple products. This changed how people worked with, used, processed, and viewed information. I like to think of this as the 0-dimensional case, in a mathematical sense, as if each user were an isolated point. 2/16

Web-pages were sources of information from companies, and this was also the beginning of #ecommerce via #Amazon and #Ebay, a kind of one-way information source with an out-of-channel messaging called a "physical product." The blogging era in its infancy was also small in comparison to the rest of the usage of the #Internet. I like to think of this as a one-dimensional vector, where information mostly flowed one way. 4/16

In the #aughts (but probably starting a couple years earlier), we saw the rise of #web2, and the mass adoption of #MySpace, #Facebook, #Twitter. Especially with the rise of #cellphones with built-in cameras and #webbrowser enabled #smartphones, the information really began to flow in two directions with #UserGeneratedContent. I like to think of this as a one-dimensional vector field, where information was flowing in two directions, but still only in one dimension. 5/16

The beginning of the #2010s brought lovely neologisms like #BigData and #DataScience to describe what companies were doing with that flood of information flowing their way. All of that data needed to be sorted, organized, and, most importantly, stored for later use. 6/16

The decade also brought the large-scale adoption of #GPU|s for the #ParallelProcessing of that ocean of data, especially with the #CUDA library and a slightly different #programming #paradigm called #DataParallel programming. This really was a game changer for data-intensive data #MachineLearning activities like #DeepLearning and started the #AIBoom we're seeing today. 7/16

All that stored data from the decade before started being used, legally or not, ethically or not, to train a wide variety of applications, some of which are actually useful instead of just being solely #Entertainment or methods of improving #SurveillenceCapitalism. 8/16

I imagine this much like a two-dimensional vector field where information per user isn't really just going in two directions, but goes in directions along another axis that describes nearly every #online activity a person has. Whether that additional dimension really reflects time, or some other property of behavior, I think is less important than the change in scale. 9/

Now we're in the middle of the #2020s, what's the next big change? To date, I haven't seen anything that really compels me to introduce another dimension or some other abuse of my #Mathematics #analogy. Predictions are really hard, especially when they're about the future, and even in this case, about the near present. 10/

Is it going to be something subtle, like a change to the color of the vectors in the vector field? I think this would be something like a change in modality, like a move to #AugmentedReality glasses, or like the recently announced #Camerabuds from Meta/Facebook or a successful version of something like the #HumaneAIPin. It may change how we interact with the #gatekeepers, but it's not a fundamental. 11/

What would it take to add another dimension to my model? This would have to be some truly #BigDeal, and if I knew what it was, I'd be talking to #VCs to get it done. However, I suspect it will have something to do with a lateral exchange in information. 12/16

Today, information flows from a user to some #gatekeeper who uses our data to profit from its use. Sure, in return, we have access to some neat-o products and cool entertainment, and some things have been truly useful, e.g., #MachineTranslation has improved immensely from access to enormous amounts of data, but it's still just a two-dimensional vector field. 13/16

As an aside, I think the adoption of the #Fediverse and the successful reimplementation of services like #Twitter, #YouTube, #Instagram, and #Reddit are fantastic advances in terms of data-ownership and privacy, but they're just changing the the direction of the vectors, not the dimensionality. 14/16

By lateral flow, I mean local interaction with our environment, or even better, with other people nearby. This has some elements of peer-to-peer, but that implies a one-to-one relationship. The big change would have to be more like a one-to-many local relationship. 15/16

I really shudder at the #privacy implications of this, but in a perfect world (maybe a federated world?) the next big leap is probably going to be a fundamental change in how we engage with our #environment. To whomever is building those products or have those ideas right now, I salute you, and wish you luck. I would also like to remind you not to be an #asshole with our data. 16/16 FIN