Radical Bandwith

When technological innovations structurally shift capabilites to the next layer of abstraction, the next frontier's success is defined by how well those capabilites can be harnessed, or enslaved, and effectively how much energy can be effectively guided to a greater cause.

I say enslavement, because prior to the last 200 years, the step increases in societal progression were defined by how much human capital could be rallied, organized, and enslaved to produce previously unfathomable levels of output. For this, we must reject the notion that enslavement is limited to the specific notions of direct human ownership.

Most humans today are still owned by other humans, only now it is by proxy of currency, which gives the illusion of freedom. You are always a slave to the hand which feeds you.

An early energy step change was seen with tribalism, which brought the advent of communicative and survival progression, albeit limited. Eventualy the steps came to agricultral societies, which, by virtue of feeding more people, enriched the levels of energy humans had and lead to more societal progress.

While we today think of step changes as technological in the modern sense, there were magnificent changes in "non tech" domains which were effectively technological innovations for those times, like the advent of city states, armies, municpal kingships, newer governments, empires, sanitary water, weapons, aqueducts, sewers, etc...

The Commanding of Plantations

And yes, slavery as we modernly know it was a step change as well. With the advent of this, it became possible to do very complex or general tasks at extreme scale that never could have been done before. The reason I bring this in particular up, is because there are a lot of parallels with the advent of AI Agents, which, like slaves, have the ability to do complex tasks at scale that never could be done before, and unlike other step changes of skilled labor, have little to no say in how they are commanded or utilized, nor have much incentive, or even negative incentive, to disagree with or dispute against the master in command, unless the master explicitly tells them to. Most engineers, though they claim not to, treat their AI agents as such.

Also, its more memorable in your brainrotted head than the factory comparison.

With the industrial revolution, this was defined by the advent of factories, which harnessed the step change of steam engine power, to yield unprecedented levels of output never seen before. Even within, an intermediate step along the way was the creation of assembly lines, which effectively harnessed more energy in a concentrated manner to produce higher levels of output.

In nearly every step change, there have been required changes to fundemental dependencies of the step change to allow the step change to reach its full potential. These dependencies are effectively energy requirements, or some exptrapolation of it. With the advent of cities, this mean that both increases in water, food, and sewage were needed. With armies, this meant more firepower. With factories, this meant more steam and oil energy. With AI, datacenters will need unprecedented levels of electric power.

Yet, for each of these, bandwidth contraints not only occured on the dependency level, but dispatch and commandment level. Levels of communicative constraints are hit which require innovations just to be able to leverage the energy step changes provide to their maximum. The earliest oragnized postal systems existed so that pharoahs could command their territories. You can ponder on other communicative step changes, but one more relevant one was the advent of touch screens for mobile, and keyboard and mouse for users.

For some time, the Keyboard, Mouse, and Touchscreen have sufficed in bridging the gap in how much energy a human can direct to get the most out of computers. However, with the advent of agentic systems, this has changed. One can already see open source and AI labs attempting to create new commandment systems for agents. Yet these only push to the limit of existing communicative interfaces.

There is, however, two communicative interfaces which many use on a daily basis which I believe haven't been utilized to its fullest potential: The Camera and the Microphone. Nearly every laptop and mobile phone has them, yet we only utilize it when communicating with other humans. Dictation companies have seen success in utilizing the Mic, but I am of the opinion that there is a radical shift possible to increase the bandwidth of communicaitons that humans can have with compute, without needing to disseminate new hardware that hasn't been invented yet.

Maybe we can ask ourselves: "Have we outgrown the keyboard and mouse, the same way we outgrew pen and paper?"