If you ever get a chance, I HIGHLY recommend talking to people.
Not just people around you *most* of the time, like your immediate or extended family, or the nosy neighbor, or the mail professional who happily delivers package upon package of online orders to your porch on what seems to be an hourly basis. (Side note: that person really deserves a Christmas gift, and I may be projecting here).
No, I mean different people, ones you've never met before, especially around the world. I have had the unique opportunity to visit many countries in my still short life, and experienced several cultures, several languages, several routines and rites. I’ve broken bread around the world, and visited hundreds of heritage sites learning about the past and present of the indigenous. I’m talking VARIETY and VOLUME here. Maximizing the number of interactions with more interesting, or less interesting people maximizes your exposure to life’s complexities and normal everyday regularities. Plus, if you happen to be in town, try to get to a Disney or 12 parks, I highly recommend Shanghai.
Based on all of my travels, and all the people I’ve met, I believe I can boil the totality of human experience down to three things: getting a cup of coffee, chatting with friends, and going home to watch TV. That’s it. It’s the same in every culture, every country, any time of year. It doesn’t matter the make, model or year of your car, your bicycle, or your footwear. If you live by yourself, with friends, family or an entire community. If you live near or in the sand, the sea, the mountains, or the jungle. It’s all the same. The sum of human experience is just those three desires.
Don’t believe me?
Consider how many times you or people you know crave a coffee, tea, or any caffeinated or carbonated beverage. Consider how many local, national and international coffee chains you see in your commute. From hipsters to ironworkers, from scientists to those experimented on, coffee is a PRIME source of your happiness.
Consider the impact of someone dropping by your cubicle to break up your day, or listening to your rants on politics, or looking at the pictures of your children from their recitals. Isn’t that significantly better than staring at your computer screen typing out a blog on AI?
Consider what you would do, in almost any circumstance just to get out of work, the traffic, the airport, the family reunion, just to get home, put on your P.J. pants, turn on the TV/open the book to watch/read whatever was on to numb the noise of the world and just laugh a little.
Exactly. Well, it’s the SAME for literally everyone in the world!
This sum of human experience is the basis for all technological innovation in the world. I mean it. Literally, every single invention or innovation exists, at least in part, to deliver the happiness from this list.
Let’s take Artificial Intelligence (AI) as just one example of technological advancement.
(As a quick side note before I continue, AI is another one of those “blurriness of definition” trigger words for technologists (like “Cloud” and ”IoT”, or “technologist”). It could mean VERY DIFFERENT things depending on the context, or the problem to be solved, or simply to the company who is either trying to sell it or implement it. For the sake of argument, lets define it going forward in this blog as the capability of computer algorithms to simulate intelligent human behavior, or thereabouts replacing any of those words with your monthly buzzword of choice. Okay? Enough said? Good).
AI is complex, and necessarily so. It’s a series of advanced mathematical and statistical algorithms applied to computer code, that can be augmented by different source data, and then re-augmented over time by the results of its own analysis. Machine learning concepts of neural networks, decision trees, random forests, regressions/classification, can be explained by professionals but MAY REQUIRE advanced degrees to fully “get it”. And that’s okay. I don’t need a medical degree to understand why I need to go to a medical doctor for a physical.
But if AI is modeling human behavior, then it must exist to augment and enhance the human experience.
Getting a cup of coffee used to be a manual experience. Get up, boil the water, add a couple of spoonful’s of Sanka and we can make a refreshing start to our day. Technology has changed all that as we strive for more accessibility and a high-quality product. The ubiquity of the internet and mobile phones allow us to use one of the hundreds of apps to locate the closest coffee chain and consume our favorite beverage. Artificial intelligence is used to learn my driving habits and recommend when I need a coffee during my 2.5-hour commute, it can predict my travel habits to know that it should pre-order my cappuccino every Monday morning at 10:17 am as I walk by an airport kiosk, and algorithms combine the knowledge of complex weather patterns, crop yields cycles, soil attributes, and field hand wearable monitors to determine the very best time to pick a plant to turn it into the perfect bean for roasting. Coffees shops are now employing robotics to make the perfect cup of joe, adding the perfect amount of sweetener and cream. And the certified Baristas’ making my Caramel Macchiato are using hundred-thousand-dollar machines that with a touch of a single button can pull a perfect espresso shot, steam the perfect amount of milk to the perfect temperature and self-clean for the very next order in such a comprehensive way to maximize the number of drinks crafted per hour.
Chatting with friends used to be a manual process. Get up, get dressed, grab my bicycle and peddle to the closest park to discuss the latest and greatest episode of Seinfeld. What a pain. Technology has changed all that as we strive for higher-quality interactions with greater distances. Now, we have the means to chat with friends, co-workers and strangers across the world. We can combine voice, video, presentations, and an almost unlimited number of participants. We can all team up and work together in massive multi-player games, and chat while we play. There are hundreds of mobile applications to find new friends, discuss topics of mutual interests, and show up at events together in a variety of homemade costumes. Artificial intelligence is even REPLACING actual friends or simple acquaintances. Bots are pretending to be our support systems, our cohorts in games, our teachers, and our initial interaction models for all forms of commerce. Robotic humans are in banks as tellers, hotels at the front desk, and concierge hosts at airports. I can chat in any language and have it automatically translated in real-time to any other language to anyone who is listening. That makes the airport taxi experience SIGNIFICANTLY BETTER in a foreign country!
Going home to watch TV used to be a manual experience. Remember the good old days when you had to walk over to the TV and manually turn the dial to one of thirteen possible channels? You picked a channel and it stayed there for the night. Luxury! Technology has changed all that as we strive for a variety of content, but precision in preference. The near thousands of channels on the television or the massive shift to “cutting the cord” in favor of online internet-based content subscription is creating more content options than one can possibly consume. Fortunately, Artificial Intelligence exists to learn my habits of watching certain programs at certain times of the day, in certain locations and can tune me into content that I (or people like me) might enjoy. In fact, AI can inject products and services WITHIN the content that I might want to buy considering it’s constantly listening in on my personal conversations, and assessing every key I type into my laptop, phone, and tablet. Additionally AI is able to actually create content from assessing 100 years of cinema and a sea of video/audio content from all the big names actors. It’s the big brother I always/never wanted! But at least I can still have a good laugh at literally the exact moment I need it.
AI is modeling human behavior and then augmenting and enhancing the human experience. But it does risk the possibility of going too far. Not in the sense of replacing humans, or humanity, but increasing risk by removing the human instinct out of decision making. The guttural capability to know right from wrong, survive the winters, and determine what’s best for life as we know it.
There is a risk of allowing AI to exist without enough TRANSPARENCY, with too much POWER, and allowing it to override our IDENTITY, collectively as the known the paradoxes of the increased value of Big Data/Artificial Intelligence.
Fortunately, we have a means to proactively deal with those human effects of automation:
Transparency: Explainable AI is a set of techniques to create a human (even if it’s only an expert human) definition of mathematical algorithms, to help assess the impact or lineage of decisions that ended up with the automated conclusion. In many ways it's helping create a window into the black-box of AI to help with walking through the manipulations of data, the categorization of information, and the trees of decision making.
Power: Assisted Intelligence will allow for AI conclusions to simply be a set of input for human decision making, however a very precise and likely highly probable conclusion. It’s the reason why complex Health-care diagnosis will still require a medical doctor, and a pilot will still be in control of an airplane for the foreseeable future for the simple reason is that only humans understand the emotional and cultural impact of life or death decision making.
Identity: Augmented AI will be a set of processes that will allow for AI algorithm conclusions to be augmented by the data “owners” to help with the learning cycle. As an example, AI might determine I like vanilla ice cream because a bunch of attributes deems me categorized that way. Unfortunately, I actually like chocolate ice cream, something mathematics can’t possibly determine. I should have a (legal) right to change those attributes or add an attribute that draws another automated conclusion. Or I should be able to choose the conclusion and AI has to learn to appreciate that choice and augment its algorithm with that choice as an attribute.
Transparency, Power, and Identity also describe the morality effect of AI automation. Where human interaction, in a non-evil world, adheres to a standard of civility and custom, artificial intelligence is operating as a digital machine, learning from a set of data to objectively deliver a result. We do know however that Data, but its very nature, is biased and decidedly so. It's impossible to define what is "good" data (both in part as a whole), as that would be subjective to every person/company and relative to any time/place. What's good for the goose may not be good for the gander. Our lack of ability to define good, ensures that all AI will learn with partially (at least) biased data, producing partially (at least) biased decisions and actions. Ethical AI, as a set of practices and techniques to increase the objectivity, or at the very least reduce the effects of bias, seeks to reduce the potential consequences and human effect.
AI is modeling human behavior and then augmenting and enhancing the human experience. It’s not just opting in our out. It’s about choosing my own destiny.
To that end, here are a few 5-10 year predictions on AI, based entirely on talking to many people around the world, and how I now perceive the totality of human experience:
Within a decade, AI will not only refine preference but will define preference. AI will be able to tell me where the best coffee is in the city, but slowly change my coffee preference over time as I try new flavors and combinations.
By 2030, AI will be able to identify emotions, and act upon them in every situation. Advertising, traffic congestion, lighting and music within commerce or work locations. Personalization becomes sensitive to my direct and continual visual and auditory range.
In only a few years, our daily lives will be enhanced with AI predefined outcomes. What path we take in hallways on or roadways, what content is consumed, and even what order we do our work. Only after we start to see the homogeneity of outcome do we insist on human augmented automation to interject our own humanity.
In only 5 years will a significant portion of content be created by AI, not in totality but certainly in repetition. Animation and CGI will be fully automated, especially in reoccurring episodic content. Deep fake content will be almost impossible to detect.
By the end of the 2020’s, AI will be our PRIMARY interface for all human-to-human communication, from determining the best method/tool of interaction, to provide real-time guidance within the discussion on effectiveness and content.
All good things, and all automation to improve the lives of the humans they serve.
I look forward to the next 10 years and what new people I get to meet.
See you out there. And let's chat.