The State of Compute Today

There is a growing snowball of complex abstractions on all layers of today’s “modern” Compute platforms.

Everything is growing in complexity, cars, phones, toys, life. Complexity attracts further complexity.

Maintainability on the other hand is in sharp decline. Just compare a car from the 1990’s vs now. Today even the car dealership is struggling to repair superficial issues.

There is a growing need and space to re-evaluate and return to simplistic approaches, values on which a solid foundation can be built. This is true for all aspects, where excessive complexity has been introduced in the past.

But how do we identify excessive complexity? Is there a universal set of laws and principles that could be employed? Perhaps.

Law #1 - Time

Time is finite. There is no mechanism to add or buy more time. Complexity attracts spent time. Conversely simplicity must result in the opposite - less time spent.

Law #2 - Value

Time spent must yield value. Value is an abstract notion, difficult to measure at times. But generally speaking, value yields as the result of a function, saving time spent on a process.

Law #3 - Use

A function is only useful when it provides substantial value. Naturally a function that provides little or no value is useless. Substantial value means the function saves exponentially more time than it takes to create and maintain its existence.

Needless complexity is in essence a violation of the 3 laws above. Its main function is consuming time without providing substantial value.

The inverted value system.

Arguably, it is not an exaggeration to state that a large part of the tech industry has failed to deliver general software efficiency. After all, this is not its primary objective, profit and customer lockin is. Everything is subordinated to maximize profitability and customer dependency. Largely for this reason alone, it became somewhat of a travesty of colorful brands and products, oftentimes offering no fundamental solution to real problems. It does however churn out solutions and products to superficial or partial problems. Successfully sells through the use of heavy marketing practices to customers who are then willing to buy these colorful products for inflated price-tags.

The entire tech space is a panopticon of this process. In some ways reminiscent of the T-shirts people wore in the movie Idiocracy, full of labels, ads and colorful coupons.

Instead of gradually advancing and improving Compute, oftentimes existing knowledge is sacrificed and thrown away, only to start over from scratch with the next fad, with another incomplete idea or broken concept.

The industry tends to reinvent its flawed products in new, sometimes even more defective ways. Products are marketed for problems which should have not existed in the first place.

Lost value

Value is lost under a mountain of low quality heavily marketed competing solutions.

The resulting fragmented product landscape drags with it a cognitive overhead which is becoming heavier by every new product release.

This phenomenon also fuels situations, where nobody understands or remembers anymore why certain things are done a certain way.

A state of “learned self helplessness” of customers trapped in the circle of repeating product iterations and reinventions flowing in an endless self propelling loop of migrations from one product to the next.

Back to basics

There must come a reduction in this cognitive overhead.

AI alone cannot solve this for humanity as 1) it lacks true intellect, 2) it is complex, 3) does not understand human cognitive processes, 4) it is marketed on the same flawed principles as any other tech product today.

There is a growing need to cut out complex patterns to leap forward. Simple stuff works.

Small incremental improvements. Secure, efficient, simple. Built on software, inline with the original ideas which lead to Unix as a system.