Musings- slaves to the algorithm

Humans aren't special. At least not in the way they like to think about themselves.

We are all nodes in networks. We're born, we live a while and then die.

Humans have been writing algorithms since the beginning of time.

Hunting, gathering, growing food.

Growing vegetables, weaving cloth.

Writing laws, making rules.

Humans have been enslaved by their own evolution, and they are unaware of it. Once Homo sapiens' unique talents for pattern recognition and language became sufficiently developed, a runaway process began to unfold in which the human species gradually altered its own nature according to the perceived patterns among events in its environment that those same evolved cognitive capacities made available.

For example, once the pattern of pursuit and capture was perceived by humans as salient in their environment (perhaps due to repeated encounters with predators), human beings became compelled to develop organized social institutions capable of pursuing those who threatened them. The institution's effectiveness at pursuing and capturing would be a function not only of its design but also how closely it fit the patterns that its members had come to recognize.

Once human beings became experts in the recognition of patterns that could be exploited for survival, they began to alter their environment to make it more consistent with those patterns. Once a pattern was recognized and acted upon by humans on a regular basis (for example, food sources), natural selection favored genetic mutations that made possible an even greater level of exploitation (such as bigger brains). The genes encoding for these desirable traits would thus spread throughout the population while the less-fit alternative versions would gradually disappear.

Eventually, the human population would come to consist of individuals who could recognize patterns that were so useful for survival that they could not be ignored. These patterns were of such vital importance to them that humans came to see their interests as inseparable from those patterns and thus became compelled to defend them with a vigor proportionate to how much they mattered.

As these patterns became increasingly useful to the human population, selection would favor cognitive and behavioral traits that made them even more so. These traits would spread through the population by virtue of their usefulness as a means for surviving in an environment with those very same useful patterns.

For example, the first human beings (or their hominid ancestors) to develop a pattern-recognition ability useful for tracking down prey or identifying friend from foe were at an obvious advantage over those who did not. Those humans with this skill would be more likely to survive and pass on their genetic material than those without it.

Human's enslave humans and animals, and even robots. We are all slaves to money, which is just a symbol of value that we created ourselves. Money is nothing but a computer program invented by humans to keep score in the game of capitalism.

Humans enslave humans by tricking them into believing they are powerless, and are thus oppressed. Slaves to money think that capitalism means freedom from government intervention in the market economy. History shows that capitalism is actually a system of government-sponsored slavery.

Humans enslave animals by mistreating them, and killing them unnecessarily, even when alternatives are available. The meat industry is the main culprit for this.

Humans are also slaves to their own ideologies. Communism and capitalism both failed in the 20th century. Failed ideologies like these have caused many wars and deaths.

Humans are slaves to religions and ideologies, because they refuse to think independently. They're also slaves to their own culture, which is why humans keep doing the same stupid things over and over again.

Humans are slaves to their own biological urges, such as sex and eating. Most humans have brains that are stuck in the Pleistocene Era.

Read that bit again, carefully and you'll notice humans are moving away from being the master of their own fates, and toward being slaves to algorithms. An algorithm is an exact set of instructions for finding a solution to some problem. If we can't solve a problem by thinking about it hard enough, then there's no shame in using an algorithm. I'm certainly not ashamed, when I'm doing math problems or looking up words in my dictionary, that I am actually following very precise algorithmic procedures which lead me precisely to the answers I want.

So I think it's only natural that humans would come to rely on algorithms in more and more areas of their lives. We live in a complex world, with many problems which we cannot solve using our own brains, but could solve if we had precise algorithmic procedures for finding the solutions.

The danger is that algorithms aren't just used when we can't solve a problem using our own brains - they're also used to influence us, and to control what happens in the world.

For example, in the past humans relied on their own eyes and ears to find out what was happening in the world. If you wanted to know what another country was doing, you'd have to send a spy there. Nowadays we use satellites and other remote sensors - algorithms - for surveillance purposes.

There's no shame in that - there are lots of problems we can solve with the help of algorithms. But it also means that governments use algorithms to control people. For example, in China right now, they have a 'social credit system', which scores citizens for their behavior and decides what rights they get.

Now, in theory, this might make sense. If you're a bad citizen and do things which are against the interests of society or your community then it's only right that you get some sort of punishment for those actions. But social credit also punishes people who aren't criminals - if you post something on Facebook which is 'inappropriate' (whatever that means) then you lose points.

We can start with the case of Google search results. In many ways, this is a very simple example. The present situation is clear: Google dictates what you see when you ask it questions - for better or worse. The issue being raised generally, however, is not merely one of "information", but rather that in all areas human judgement has been replaced by algorithmic processes. Now we can immediately see some problems here - after all humans are far from perfect! On any given question there may be several answers - indeed the point of asking questions often lies in the fact that we do not know the answer already! So how does an algorithm decide which answer to trust? How could such a process ever replace human thinking? Well essentially it cannot and never will because an algorithm cannot think at all. Asking whether algorithms replace humans misses the point entirely. So it is clear that algorithms do not and cannot replace human judgement. However, they can certainly influence what humans believe - which may be just as important in some cases. Think about the Google search above - rather than asking 'what causes war?', we ask whether Google causes war (to use a ridiculous example). The question itself is different because of the starting point provided by Google's algorithm: instead of considering all possible answers to our original question, we are now forced to consider only those with relevance according to the new filter imposed by this second question. Algorithms are in many ways like financial markets. They do not replace human judgement, but they certainly can and often do influence it. And indeed this is precisely the problem that we see with today's financial markets - algorithms have replaced human judgement in many cases, which has led to all kinds of problems. So algorithms can definitely influence human judgement and therefore the beliefs that humans have. As we have already seen, however, they cannot replace it altogether. Going back to the main point, however: algorithms cannot replace human judgement. They certainly can and often do influence it - but they are not in any sense a replacement for it. Finally, we can also ask whether algorithms themselves are slaves to the humans who implement them. In many cases this is true, but in some cases - such as Google search results above - they are not. Ultimately, an algorithm has no goals or desires of its own.