Analysts Suggest We May Have Reached ‘Peak Algorithm’

Nowadays everyone at least has a passing understanding of what an algorithm is – it’s the clever bit of tech wizardry that’s employed by seemingly every app or platform we interact with. Their uses vary, but at root it’s all about filtering and customizing information for individual users. It’s easy to forget that this whole thing is still a relatively recent phenomenon.

Before 2006, the word was seldom heard outside of the realms of science, engineering and mathematics but it began to steadily gain ground and prominence as Web 2.0 – the age of social media, came into view.

You may wonder why we even need these things – but if you pause to consider just how much data there is online today, their utility becomes obvious. Algorithms sort through this data in a way that enables you to more easily access the content relevant to you.

There are a whole host of applications for this. For example, if you’re a fan of horror movies on Netflix, this streaming service will quickly learn this preference and make sure you know about new or popular horror films that hit the platform.

Algorithms can also be used to make product suggestions when shopping, drawing on previous market data to make predictions. This is how a website like Amazon knows that if you’re buying an inflatable mattress, you may also be interested in picking up an air pump to inflate it.

So far so good. In fact, it raises the question as to whether there’s any issue with all this at all. And in truth there’s some nuance there. Algorithms are not ‘going away’ – they’re an indispensable part of how people use the internet today, but their deficiencies are better known now and certain platforms and services are looking for alternatives where it seems more suitable.

To get to the heart of this, we need to understand what possible critiques people have towards algorithms. One is that they can become easily ‘over-trained’. We’ve all had the experience of watching one too many cat videos on Instagram, only to suddenly find your feed has pivoted to a seamless tapestry of cat content.

Algorithms can also create so-called echo chambers, where you run the risk of missing out on all aspects of a debate or news story due to your intrinsic biases that the algorithm then amplifies. Finally, to work well these tools need large data-sets. That makes them a poor fit for certain services where they cannot leverage such information to make informed predictions. So what are the alternatives?

Niche Curation

Algorithms are intrinsically ‘low trust’, given that they’re the product of little more than number crunching. That makes them a poor fit for providing recommendations in a sector where human input, expertise and know-how are considered valuable. In light of this, there’s a growing number of platforms emerging focused on providing users high-trust, human backed recommendations for a given goods sector, or service.

For example, CasinoReviews has established a strong standing in the iGaming space for providing comprehensive breakdowns and rankings of quality online casinos. In addition to this, it is empowered to furnish users with competitive welcome offers and bonuses for these platforms, further engendering a trustworthy alternative to algorithmic suggestions, and one that is cost effective to boot.

Escaping the Echo Chamber

Sometimes they say if you can’t beat them, join them. That’s precisely the approach being taken by a new generation of platforms that are looking to embrace the intrinsic biases of algorithms to gain a broader overview.

One prescient example of this in action is the news platform GroundNews. This website aims to provide its users with an overview of news coverage from across the political spectrum. Knowing full well that the modern social media environment emboldens confirmation bias, this service furnishes users with a spectrum of outlets covering the same story and empowers the user to make up their own mind on today’s issues while being fully apprised of the leanings of the individual sources.

While this is seen as a frontrunner to this approach, there’s no reason why this kind of corrective measure cannot be employed for other forms of content. For example, Spotify – with its sophisticated map of user music preferences, it’s perfectly empowered to provide you with a playlist of songs and artists that run counter to your existing interests. You may know, your new favorite band or artist may be looking on the other side of the algorithm.