My Personal Experiments With AI, Chatbots, etc
Unless you’ve been sleeping under a rock, you would know that 2016 was the year of the bots. Tay happened, where ‘real world people’ taught an artificial app some ‘real world’ abuse and racism.
Around that time, I started to experiment with bots for my work and eventually started experimenting for myself. For millennials out there — if you stop reading here, I don’t blame you since I am writing about your world. However for someone in my generation, this is the proverbial, brave new world. I’ve worked through about 40 AI apps, a couple of dozen chatbots for Twitter, Facebook, LinkedIn, Tinder, etc — below are my honest findings via the experiments on the top four platforms:
#1. Experiments With Twitter. There are so many bots out there — Olivia Taters, Dear Assistant, AutoTWBot, robolike, FriendlyBot — the list is endless. You can create your own bot, as well — in less than 30 minutes.
For example, if ABC sat on my Twitter liking, retweeting and posting based on my settings. Of course, I didn’t leave ABC without adult supervision. Every now and then, I did a surprise check on it and I had just a handful of followers, plus, I wasn’t important enough in the digital world, so a “Tay-like” disaster wasn’t on the horizon.
Overall, the AI bot performed well and in a week my twitter following tripled. If someone asked me, were those interactions real? Are those followers people? Frankly, I don’t know — I was just pretty smug with the numbers.
Though, every once in a while the cat was out of the bag… Once, this cute little Chatbot of mine, decided to follow a 20 year-old, from Oklahoma — and here I am, almost 40 and a Senior business leader based in Singapore. The surprised lad messaged me back asking why I followed him. Seriously, WTF was my bot thinking?! I quietly unfollowed him, laughed to myself and thankfully, no harm was done.
Imagine if what research says is true — if really, 15% of twitter users are bots. Having said that, I am waiting for #SAP procurement bot. We’ll see…
#2. Experiments With LinkedIn. Emboldened, I used a similar app for LinkedIn which filters out content and posts it — with some posts automated from twitter. Again, there are tons of apps available — Drum Up, Followinglike, Linkedeep, Profile Hopper, Devumi, etc. Professional networking is a viewed as a task, where everyone is out for something — a better job, sales, network growth, prestige — something.
Anyway, LinkedIn interaction metric zoomed, SSI score increased by 48 points in a month, and surprisingly, it didn’t take much of my time. I did use those interactions to connect with many people in my professional circle, but social media became a new game for me. Checking scores, working on metrics — I was running a geeky fantasy premier league, entirely of my own creation.
In these otherwise perfect interactions, there were occasional misses. Messages like — “Are you real? Am I talking to a robot?”. I would scroll through the conversation, and yes, it did look like my chatbot was zoning out. Parental supervision back in action. I messaged back with a few typos, confirming that I was a real person. (Thank God for parents.)
#3. Experiments With Facebook. My experiments on Facebook didn’t go really far. Possibly, because I have many “real” and old friends I like to interact with them every now and then. I set up Facebook messenger to auto-reply to some recruitment mails I receive, but nothing very interesting happened.
#4. Experiments With Tinder. I hate flirting. I am super bad at one liners, and detest small talk, as a matter of principle. Fed up with friends pushing me on dates and emboldened by the experiments above, I installed an auto liker for Tinder. Tinder bots have pretty chauvinistic names like Casanova — but there is also, Ghostbot, Tinderbox and Bernie.
Initially, one needs to make about 50 choices, provide adequate data for the app to run on its own. Afterwards, the bot goes about right-swiping for you — taking the hard work out of flirting. Casanova auto likes 100 “girls” in your area, sends an opening line based on their personality type and continues to “chat” until the phone number is acquired. Of course, the 3 day wait before you call or not, is entirely your decision, but there’s a good chance you will never know if they’re hot or bot — until you meet.
Imagine waking up to dates auto-synced in your calendar. Or, on the way to your date, the bot messages, “She likes yellow flowers” and beeps at the restaurant, “She likes Hendricks…allergic to avocados…”. Killer app right? ?? Maybe she dates a bot, as well.
Most likely she knows nothing about you either and we are back to a blind date scenario where discovery commences on the first date. All awkward conversations are back in the real world, sitting across a real table, eye contact — and the World is back to square one. ??
I am not one to talk of the ‘rise of machines’ or the robotic take over of the World kind of stuff — but, this is fascinating. Technology has given us the World we live in — Let’s see where does it take us and thankfully, we are all in this together.
P.S. All apps have been removed from my devices, as I write this post… I enjoyed the game — while it lasted.
Anu
Twitter: @anulall
I help burnt-out, anxious, over-achievers transform to ? Fit, Strong & Happy ?? With 3,000 year old mind-body techniques. WITHOUT twisting like a pretzel, eating leaves or meditating for 8 hours ??
7 年Thanks Rashmi Deepak AI is the industrial revolution of our times... and makes so much sense aswell. Leave the human intellect to more meaningful interactions. (Not from a bot :)
Head of Sales Enablement, Asia Pacific at Cushman & Wakefield | Strategic Marketing & Sales Growth Strategy
7 年Excellent insights Anu, super ???? and this is not from my 'bot'!