There will be a lot of people who will click on this article, look at the title and think I'm crazy. AI seems to be everywhere in the news lately, from the massive data crunchers employed by Facebook and Google to find out what cereal you're eating to the scary facial recognition systems that could transform the way we think about crime prevention (for better or worse). There's so much attention on the big flashy developments that no focus is being made on the projects that are actually making a big, tangible difference in people's lives now.

I' reminded of a Vox article I read a year ago. It was about a Twitter user who made an account titled: 'on Mice' and sought to debunk every sensational healthcare development by pointing out that the authors of the clickbait news were omitting a major feature of their revolutionary new procedure or medicine: that they weren't actually tested on humans, they were tested on mice. I don't want to discount the effect of a successful trial on mice - it is a major step in the right direction, but omitting that the results of an experiment were only successful on a test subject like a mouse hides the fact that any actual effect that the discovery will have on human life is far, far away.

AI is in a similar position now. Every morning, when I go through my gauntlet of tech-related news articles, and I see the newest method for predicting protein folds, or attracting customers, and so on and so on - I have to remind myself that none of this was written by someone with the same mindset as the researchers who produce the white papers I read everyday at work.

This contrast came up when I was describing how the use of a new tool - Google's developments in a method called Neural Structured Learning that uses metadata from graphs to improve machine learning models in scenarios with little labeled data - would improve our model's accuracy by 5-10%. I was excited - my boss was less so. 'Five to ten percent?' He asked. 'Why's that significant? Why should I devote resources to looking into this?'

That question highlights my point - the AI revolution is subtle and coming quietly. Too many investors, C-suite officials, hobbyists, and pundits are enamored with the glitzy promises of AI spoken about at TED Talks, tech conferences, and espoused by 'industry leaders' like Elon Musk. AI - or the mechanics behind it - do have the potential to introduce life changing effects in our lives. The potential of having an analysis system able to pick up on nuanced patterns in our world and make use of them to produce unimaginable outcomes is a really cool thing. However, this talk obscures what's actually going on behind the scenes.

All of those technologies are either 30-40 years from being condensed into a product or application that can actually be used by interested parties. I'll go into the specifics of this claim in a bit, but the main gist is that these pundits are describing them as if they'll be unleashed tomorrow. As if the data that's used to train them is already perfectly formatted and the hyperparameters and objective functions of the models are able to generalize far beyond a specific use case. This is not the case. Even if you manage to get a working machine learning model that can extrapolate from a good set of data, you'll often find that its actual application to human lives and behavior is something that hasn't even been considered. Sure you can produce more data from it - but how do you actually use that data?

That's the reason that I brought up the comparison to medicine. Expecting revolutionary AI to become a part of daily life is like seeing that HIV has been eradicated in mice, or that a pill can prevent allergies, and expecting those solutions to be on the market tomorrow. Decades of time, billions of dollars, and countless minute changes built through exhaustive human effort will have to go into those solutions before they become a 'revolution'.

I've focused on the negatives of this kind of AI talk for the majority of this article - so let me take time to expand on what is a reality - now - and how you, as a hobbyist, a C-suite official, an investor, and so on - can realize the contemporary potential of AI and stop focusing on all the noise that's distracting you from what's actually valuable right now.

There's nothing particularly flashy about creating a new data filing system for large hospital systems, or giving an aspiring architect the tools to make their resume stand out, or improving supply chain efficiency by 2%. But these are the features that are actually changing lives now. Instead of trying to build some flashy new processor that can analyze millions of facets of data at once, most businesses should be looking at the wealth of tools out there and seeing how they can integrate them with their business model.

Tepid responses to unexciting AI-related tools are obscuring the actual, tangible value of what's available right now. To modernize an old proverb, the internet wasn't built in a day. Social media didn't become the influential juggernaut of marketing over the course of a year. What happened was that small, medium, and large businesses found a small feature, a random off-the-shelf model, that could improve one small aspect of their business. These gradual improvements were the reasons for promotions, pay raises, amazing returns on investment, and the foundation of the careers of the folks in charge today.

The AI 'revolution' is, and will continue to be, just as quiet. Your life won't change over the course of a single 24 hour period due to AI (yet). What will make a difference in your life - no matter what field you're in - will be your adoption of a simple programming creed: if you do something repetitive 3 times, there's a way to automate it. This kind of thinking will allow professionals of every speciality to acknowledge that there is a place for improvement in their field.

Business that have an edge on their competition right now are the ones adopting this mentality. They're the ones who are ignoring the flashy, attention-grabbing clickbait that won't actually produce returns in favor of the smaller steps of automation and expedition that will improve their bottom line. Developments in AI today will be an industry-warping, life-changing feature of the 21st century - it just isn't happening where you're looking.

I really dislike reading articles that just offer one author's perception of a trend and don't actually contribute any meaningful courses of action to improve a situation. I'm a policymaker - I like advising on what some entity can do to take advantage or improve their response to a trend or scenario. I want to end my article with a defined set of prescriptions to follow to avoid being captivated by the headlines and actually use AI for something useful, now.

Here's how you can begin to use AI today:

• If you have an 'innovation' staff or staffer, whose clearly defined responsibility area is to identify off-the-shelf open source technologies that can be used to help your objectives, have them check in with this list of resources at least once a week.
  • The MIT Technology Review
  • Papers With Code
  • Skynet Today
  • Open AI
• If you don't have an 'innovation' staffer whose sole responsibility it is to intake information about your company and its product and figure out ways to make it better - hire one!
• Have your technical team (hopefully with backgrounds in tech and a willingness to learn new things), go through Google's tutorials. Google is a gold mine for any company that wants to begin improving their bottom line with developments in machine learning. Their products are specifically designed to be expanded upon by companies and it's ALL FREE. Here are a couple links to get you started:
  • Google AI Tools
  • Open AI Gym

All of these tools are specifically designed for creative implementation. This is how AI will revolutionize the world: not with a bang, but with a small, sustained push for innovation.