AI is a Bubble & Everyone is bad at Data Science.
There are BRILLIANT Data Scientists, Machine Learning Engineers, and AI researchers (gonna just smash it all in and call it “AI” to make it easier, k? Cool) all over the world. They do amazing things. Today Data Science and AI working together can understand everything about any person, anticipate them, and even change how they behave. Wanna know why your group can’t do that as a business with all your customers? Why they drop off on entry or only small amounts of them ever pay money and you don't know why? You.
People stop people from understanding people.
You see, this happens with just about every major shift in “innovation”. We are often defined in character by the practices of our work, and our work is always dictated by the exposure of our early development, for better or worse. I won’t go into the intricacies of human development or cognition, but in so many ways they directly correlate to AI and Data Science. More importantly, the state of them today. Buy me a drink some time and we can talk shop.
In the days of Brick and Mortar, it was about relationships or those who could manipulate them. Naturally, there was a balance of things that resulted in various businesses and practices. At the time it was hard to determine who was who, separation of people and information leaves a divide of knowledge and by extension a lack of growth socially, and interpersonally. Then software came and It changed how businesses did everything. The groups that couldn’t migrate from a register and ledgers to a digital POS or file/inventory management were hurt, ESPECIALLY those with growth. And the only thing that stopped them was people who made decisions.
Then the internet. By now, massive Brick and Mortar companies who were raised with software, or at least understood it created the Blockbusters of the world, marrying the two. This was about global expansion and “BI” driven functions. The market grew, but the mentality of leaders looked at the internet as “R & D” or saw the short-sighted approach of “Sell more physical goods”. They didn’t consider that the market's values and consumption habits were changing. Netflix devoured them. And the only thing that could have stopped them was the people who made decisions.
Now we are in an internet age, knowing what the internet is and how it works is just part of the job. You basically cannot exist in any market today without understanding it. Solutions are being made all over the place to make it easier to use, market yourself, market businesses. Heck, I am using it now to make everyone awkward calling out how amazing people are at being bad at things, while also letting you know its ok, and it's human. You’re doing your best, we all are. Also, we now have amazing black boxes of AI and Machine Learning and Blockchain! Except no, you sorta don’t.
Hey Bruh, we‘re “Data-Driven”
Nah, you probably aren’t. You probably trusted someone and hired them to tell you what you needed to hear to feel you are. Or that person goes to conventions where someone spins how their tech and tools help you do that. There is a logical correlation to the growth and attention of executive leadership and the individuals they align themselves with. There are fewer “types” of company cultures now, there will be far less soon, I imagine.
The reality is, you maybe have a very small amount of data you actually understand. It’s true you have a lot of growth, but if there are parts of your business you’re facing pains with, that is just one perspective of data you are lacking. And there are lots of pains. And like all things, people stop this from happening. But it’s ok, these are growing pains, some must fail so others can learn and succeed and you’re training the next generation of “Entrepreneur” of things they should never do. So thank you in advance for the lessons for our youth.
It's a foundational topic, with abstractions of issues along the way. Foundationally, you have the people that manage the Data, and build for it. I am gonna guess most, if not all of your groups basically put them on the back burner. Characteristically, these types of minds are not outspoken, they are trained out of academia and demeanor of colleagues to follow the same approach. Their cognitive bandwidth is reserved for complex topics. They also shouldn’t and don’t have to think about how data is being used by people who need to understand it. So, they collect all the things. Good luck and have fun (#GLHF?)finding anything there.
For funsies let’s say all Data is “Legos”, it's like if you wanted to know exactly the shape and color of certain legos you have, and all the engineer really cares about doing, or should care about is giving you a bucket for em all. Then one day you want a gray one with two notches on it. And we have a LOT of legos with big businesses, you're gonna have to find that. This is also assuming you really ever want to find ONE Lego type to represent one piece of data. Guess what. Nope. And if you want more accuracy, you gotta have more types of legos to connect, then you gotta connect them over a sequence of time, then you gotta do it while other people are dumping more legos in. But hey. AI? Wait. Science. Enjoy.
Then, we have the “creatives”. Not to say some are not incredibly calculating and lean into data. But very few creatives design products in any way that communicates to data engineers what is being tracked. Mr. Boss Person, so it's clear, how Data Teams architect data has EVERYTHING to do with how effective your Data Science, Analyst, BI, Unicorns, whatever, can operate. Ultimately any “Data Buzz Word” team is doing work against an ocean of “WTF does this even mean”?
Back at the lego thing, the shell of something really amazing is made, like a castle so you can see surface-level legos and how to find them because it was designed that way. But all the stuff on the inside that really makes it work, or whenever your lego buddies live, or how it lights up, where things are moving around, was never structured in any way, so you still gotta go digging through that rubble of legos to make sense of how that structure is even standing (your product or business) if your lucky its cool. Or, someone finds that busted structure and makes a better castle, or people that built YOUR castle are fed up and just wanna build their own because that's crazy talk trying to fix that thing. And the Data folks know the only way to really fix the problems is changing the castle as a whole, but good luck pushing that agenda.
Don’t get mad at them, it's not their fault, you never gave them a seat at the table. And honestly, its totally understandable why any group does it this way, we are very human, we are not machines, but this is certainly why we need them. I have explained what “Data Science” was in the past, so I won’t go into it here. Suffice to say, all that crazy AI, and ML, and lakes, and pools and oceans of data are typically there as a result of cultural authorities you put in place, who are making up for creative intuition that results in chaos in data, which makes it harder for any data group to navigate, actually, near impossible, and harder for any decision-makers to react on data without a majority of the call being intuition from their perspective at that time. Which could have changed over an article over coffee (which I am hoping might be this one). Sounding familiar? Blockbuster? Netflix? Eh?
But we use AI everything!
Yeeeeaaa, also not really. I mean you have it, but it's not really doing much. IDC what industry you’re in, or how big your company is. Ultimately, any AI plugin thingy, assuming it's actually what I would personally consider “AI”, can only give you any real impact with things you already comprehend as a company. So, basically, you’re just milking the things you found by intuition in the first place. You’re not really changing much about you as a company. You got fed something about how some specific number you were trained to think was important is now doing more of something with that number. Dope.
What you didn’t understand is that there was a systemic impact of that practice which was leading to pain somewhere else. Then you sent analysts who just sorta deal with what they have and have a practice of just knowing its gonna be taken at face value based on who sees it. Or you're gonna have a certain kind of “Data Science” person who is gonna do who knows what to what kind of data based on what they have access to. And nobody is going to actually articulate what the suits REALLY wanted from all of this.
As a result, confidence will be lost in AI. More analysts and BI folks will organically become Data Scientists because companies need that to save face, all the while very few will ever actually do the prediction of anything meaningful. You have a very comfortable executive somewhere that already built out all this legacy and you sorta just gotta do your best, then after an army of Data folks who do a thing, if they ever find a candidate or leader who ACTUALLY knows Data Science, no way that person is getting hired or that suit is getting rolled on, and they know it. Or worse still, they just don’t understand what is being said to them at all. Yipes.
AI is already outpacing you and probably always will
There are a few groups out there that are making a transition to being AI companies now. It’s “funny” because of their transition to doing so is leading to some scares in stock prices in the tech sectors. If it’s pains in the Digital Entertainment and Games space? Nah, they are just really bad with data. Fun fact, you can often tell how slow intuition/emotion driven by the amount of data roles you have in any group. BI, Analyst, Data Science, User Research, Market Research, it means there is a divide in data location, practice, and reporting. So, HOORAY MORE GUESSING! Sensitive topic. Can you tell? Oi.
Protip: If you go back and check the origin and function of each of those research and data organizations, you can basically eliminate the role, promote them into Product and Design teams, migrate their backgrounds in standards and workflow, have cleaner data, better use of AI, do more without bloating the team further, have fewer questions, and actually be data-driven. All you did by having each of these groups is take something from the 70s, 80s, 90s, and 00s and smash it into a massive political structure. Time for some innovation in people. ALL the people, including me.
When AI or Data Science can act on their own, they are capable of amazing things. Finding world-changing truths. Data Science as a practice in Academia has always been a thing, it never needed a title. AI was naturally a path of growth from various fields of quantifiable academic practices. AI is looking for more and more data to do things, all the images, and cars, and energy consumption. It's all Data. That’s it. The groups with less to deal with in scale have less risk of staying on top, simply given the direct correlation to “truth” and “leadership” from bottom up. Innovation and growth is their culture, and it happens because they are in positions to understand people.
I don’t think anyone right now truly is the next “AI” company. The next “Google”, as it were. But I do feel strongly that Microsoft, Google, Electronic Arts, Activision, *insert user-facing technology entity with all the monies* are at risk of being consumed by the next true AI company. And I suspect while those groups may try and be acquire the first big “AI Company”, it's more than likely the individuals leading it don’t see validation and growth the same way most of us do or define success by the currency of money. To these new leaders, Data is factually their currency, it's not just a tokened phrase to feel relevant. As a matter of fact, they are very likely the ones we dismiss today playing and interacting online, hiding among them. What we know about topics of scale today, is that there will be far fewer amazing things in the world among the masses, but those far fewer things will be far more amazing than we can appreciate now.
We don’t even speak their language, these new AI leaders. Right now, all we are doing is laying down a foundation for them to learn from our pains and successes. They are trained with transparency, and we call it “Catharsis”. They consume everything and find cracks in our systems, and we call them “Hackers”. They literally rally to call pains out for things they are passionate about and we call them “Entitled”. They care very little for what people have to say about them and continue to grow while others debate the feelings of it all, and we call them “Psychopaths”. They will pick up those pieces we left behind and do something more with it, while we potentially create this bubble that is AI.
We are all subject to our analysis of our snapshot in time. Our perspective inherently blurs our perspective of what things are. It's so comfortable and so human to try and rationalize any moment in the society of a relationship with a thing, classify it, label it, categorize it and accept it based on the snapshot analysis of society. Factually, everything we think about AI, how we use it, how we work with it, what it means to us as a society, will change. And for all the things we say of our younger generations, all of those labels, the thing that stops us from using, or understanding AI to make an actual impact, is “Ego”.
