Attention (as a service) is All You Need

Going “all the way back” to 2017, to Google’s groundbreaking paper titled Attention is All You Need where we got first introduced into a novel neural network architecture known as the Transformer. This architecture revolutionized the field of natural language processing (NLP) by introducing the concept of self-attention mechanisms.

Today, AI is way beyond voice assistants in our homes or recommendation algorithms on streaming platforms. It’s insane and gets insaner by the day, but the best part is that it’s out there for everyone to use.

About services

A service is a digital offering that allows software applications to interact with and access specific functionalities or data provided by a remote server or service provider over the internet.

It’s something, somewhere, that we can ask to perform tasks on our behalf and expect a response. The web, as the vast ecosystem that it is, cannot function without this.

Most (if not all) products today rely on third-party integrations. If we’re building an app that includes weather forecasts, we are most likely not creating an entire forecasting product from scratch; instead, we will use a service that provides answers when we ask the right questions.

Well, it turns out that a significant chunk of the AI breakthroughs achieved to date is available in this manner. Most of them come at a modest cost, but the fact that anyone can utilize these revolutionary breakthroughs without breaking the bank is truly amazing.

What does this mean

The world is populated by all sorts of people, including those who invent cars, airplanes, or the internet, and those who build brands and services on top of these inventions. We will always need both types of innovators as they collectively propel us forward.

This means that to create a jaw-dropping product, enter a new market, or revolutionize an industry with AI, you don’t need a Ph.D. in Mathematics. All you need is to leverage the available AI services and, of course, an idea. The latter can be the challenging part because we are just beginning to understand the possibilities of this newfound power, and as soon as we start grasping it, it advances ten steps further.

Today, it may seem obvious that we have social media, online banking, or ten messaging apps, but we have fully embraced the era of the web and apps. We are already deeply immersed in a whole new paradigm, and most of our minds have not yet fully explored its creative potential.

Let’s not forget that the invention of bread and the sandwich had a gap of thousands of years.

Mass adoption

Unlike blockchain (which remains an amazing and largely untapped technology), there is no “if” in mass adoption when it comes to AI, it’s already here for everyone to use. It’s already revolutionizing well-known products like Photoshop with its newly discovered image editing capabilities, and let’s not forget that AI Has Already Created As Many Images As Photographers Have Taken in 150 Years.

Take the example of invideo with it’s new release where you can have a complete product promo video done with AI text prompts. Although VERY impressive and challenging of a product to create, I’m sure, it’s not “2022-challenging”, and everyone else needs to keep up now.

Long story short, it’s here and it’s creating entire new markets.

Final thoughts

It will take some time before AI becomes ubiquitous, but not nearly as much time as we are accustomed to. The tangible and significant advantages it offers to businesses will drive its adoption forward like a freight train. Innovation never comes without risks, considerations, or regulations, but those are topics for another discussion.

In the near future, we will replace “There’s an app for that” with “There’s an AI for that,” and I can’t wait.