The Magic of AI

As a child growing up in small and remote mining villages in the 70s, I remember the excitement and sense of importance I felt when my mother allowed me to wind the crank phone before she lifted the handset to speak to the operator. It was an era that lent itself to visiting friends for a chat, and writing letters to friends and relatives that didn’t live close enough for regular visits. Unless it was necessary to call the local doctor or make an appointment with the hairdresser in the nearest town, there wasn’t an awful lot of switchboard activity. This meant that the switchboard operator’s day could be rather boring and other than answering “nommer asseblief” and plugging the ringing cable into the relevant jack, downright soul-destroying if it weren’t for the delight of eavesdropping. Yes, Generation Y, before Google we gleaned all the gossip  from our village switchboard operator!

before artificial intelligence a crank phone

It is, therefore, no surprise that I am impressed by artificial intelligence every single day of my life. I marvel at the convenience of interactions with Alexa, primarily because she doesn’t become exasperated when I ask questions about things I don’t understand.  She doesn’t seem to mind being asked to turn the lights out, and she doesn’t forget to tell me when to take the chicken out the oven.  As you may have gathered, being a child in the 70s makes me a quinquagenarian and old enough to have adult children who, it seems, have forgotten that I answered all their questions (repeatedly), turned their lights out for them, and reminded them when they needed to do things.  Just saying.

By the same measure, this means that I have witnessed the evolution from crank phone to smart phone – as if by magic, for I was distracted by life’s sleight-of-hand and not paying close attention, my telephone is now a tool I use to access and benefit from artificial intelligence.

For all the benefits, however, there is a concern that my phone, designed to facilitate the collection of data that AI uses to perceive information and initiate actions –  location history, banking, and a great deal more, stored in the cloud – is potentially vulnerable to data theft and even possible exposure.  It is terrifying to think of the privacy disadvantages, but even more so to imagine living without the convenience of a connected life. And, so it is, that I have opted to risk my privacy in favour of a talking map that tells me exactly when to turn, veer right, or take an offramp.

Impressive

Artificial Intelligence I like

To clarify, I’m impressed by the daily conveniences of AI.  I wake up in the morning and use my fingerprint to access my phone.  My fingerprint!  That’s incredible, and I can honestly say that as a young child imagining what the future might bring, never once did I consider such a far-fetched notion as fingerprint access to a smart phone.  The weather forecast is my first check for the day.  I’m at an age where the weather is a large part of my social conversation and knowing the predicted pattern for the day or, indeed, the week, provides me with the information required for polite small talk, something I abhor but have found necessary if I’m to maintain a professional network.

Secondly, I check the load-shedding schedule – another vital snippet of information that allows me to plan my day, and bulk up polite-conversation content. This activity also gives me the opportunity to tut-tut and shake my head in disbelief at the government’s ineptitude and the irony of AI reporting on the failure of natural intelligence.

The third AI phase of my morning ritual is the completion of Wordle and a crossword, courtesy of the New York Times. I drink my tea, smoke a cigarette, and solve the puzzles. I don’t need to walk out and dislodge a rain-soaked newspaper from the gate, or hunt for a pen.  The crossword doesn’t become a messy inkblot of corrected letters over bad first guesses as I reconsider, backspace, and retype.  It’s neat. It’s enjoyable. It’s convenient.

My workday, a flurry of SEO and marketing, relies on AI for performance reports and analysis.  Facts and figures for copywriting and interesting information for content are a mouse click away, and I can honestly say that I don’t remember how I managed before Google.  I do remember a magnificent set of Encyclopaedia Brittanica that lined the bookshelf when I was a child, which we were permitted to read if our hands were clean.  How difficult it would be to do my job today if that were my only source of reference!

Throughout the day I’m distracted by personalised social media feeds and Google adverts, both based on my search history.  I don’t really need another pair of jeans that promise to lift my butt and flatten my stomach – I have several already, and they do not make any improvement to my body shape – but AI is designed to look for opportunities to sell me more of anything I don’t really need.  And, based on my viewing history, Netflix let’s me know when there’s a new series I might enjoy – long gone are the days of checking the TV Guide in the YOU magazine for the viewing schedule!

Google Maps is, in my book, one of the best and most convenient examples of AI.   I never worry about getting lost, running late, or traffic jams.  Google directs me to alternate routes to avoid gridlock and re-routes me when I do take a wrong turn.  How very different from family road trips in the 70s, and the amount of time my father spent studying maps prior to departure! Not often, but every now and again, a new road or a wrong turn meant we got lost.  Whilst my father cursed under his breath and called anyone driving slower than him a moron, and anyone driving faster than him a maniac, my mother would spread the map out on her lap and try to give my father directions. My father’s mood would worsen when he noticed my mother was looking at the map upside down, and my mother’s patience with him would wear thin when he refused to stop and ask for directions.  And all the while, after hours in the backseat, three bored children whining about being breathed on / pinched  / looked at by their siblings.

man stranded without artificial intelligence

I must admit that I feel a little lost without my phone and despair that, perhaps, I’m losing basic skills in the absence of practice.  These everyday AI conveniences free my mind of small tasks so that I can focus on bigger tasks, but at what cost?  Would I be able to navigate from one address to another without being told where to turn and using only my dulled sense of direction?  Probably not but, then again, does it matter that my mind is not as smart as my phone?  Well yes, I think it does if only because my reliance on AI clearly limits my autonomy and my ability to innovate.

A Little Less Impressive

Artificial Intelligence I don’t like

For all the convenience, however, there is one instance I would much prefer was not managed by AI:  customer service.

Chatbots, intended to automate the most repetitive and redundant customer support enquiries, have proved to be a source of increasing frustration.  A survey conducted by UJET (a cloud-based call centre app) reported that 72% of the respondents considered interacting with automated chatbots a “complete waste of time”.

As businesses cut costs by eliminating live customer service staff, overall customer experience declines.  78% of consumers are forced to connect with a human after failing to resolve their needs through an chatbot, and it is further reported that more than half of the consumers that took part in the survey believe that a phone call with a live agent provides the speediest resolution and best customer experience.

This is an interesting conundrum, if ever there was one.  Chatbots provide the perfect customer experience, in theory:  pleasant (they never have bad days), fast, and polite.  They are, however, unsatisfying; and, yet the move towards full automation is on the rise.  Automation promises a cost reduction of over 30%, but is that significant if customers go elsewhere for customer service that isn’t frustrating?

Customer centricity demands that the customer is the focal point of all decisions related to delivery products, services, and overall customer experience and satisfaction.  Businesses spend a small fortune (quite possibly some of the savings they’ve accumulated from automation) focusing on customer centricity, sending their remaining live employees to workshops to learn about the subject, and aligning their business strategies to incorporate the concept.

AI chatbots to save money

Now, being customer centric means anticipating the customer’s wants, needs, and communication preferences.  And then, getting it right.  Surely, in a world of brilliant tech and business minds, is it not by this time evident that while AI is nothing short of amazing, it is not ready to replace humans? 

15 years ago, I could pick up the phone and call my banker.  She knew my name, she knew my business, and she provided customer centric service.  I didn’t bank with the institution because it was reputed to be the best, but because my banker made every effort to demonstrate that she valued my business and made every effort to retain it.  Yes, I know that times have changed, and while I’m all for progress, progress that results in inconvenience, frustration, and downright dreadful customer service is not a customer centric solution.

This begs the question:  who is not thinking?  Is big business so taken with cost saving and large bonuses based on the bottom line, and so secure in the knowledge that it is the same the world over, that customer centric solutions are a thing of the past, and nothing more than meaningless words incorporated in their value statement?  The idiom  “If it ain’t broke, don’t fix it” springs to mind (well, the little functioning bit I have left, anyway), and I wonder if, in years to come, AI will evolve to match human capabilities like reasoning and complex decision-making, or if people’s increased dependency on machine-driven networks will erode their abilities to think for themselves, function independently without automated systems, or interact effectively with others.

 

Very Impressive… but, oi, the implications!

Artificial Intelligence without ethics

DALL-E is a machine-learning model created by OpenAI that produces images from language descriptions.  It is a neural network algorithm that generates a photo or a painting from short text descriptions of a scene and, even though I have no idea what a neural network algorithm is, using DALL-E is a great deal of fun! I typed the following text “orangutan on a wooden bridge in a jungle” and DALL-E generated this image in all of 20 seconds.

 

Artificial Intelligence Dalle painting orangatun

Disheartening, though.  I haven’t worked as an artist for almost 10 years, finding the prospect of irregular income less appealing with age, but I still paint the odd picture when asked to.  My daughter has a wonderfully eclectic home and, about seven years ago, sent me a reference pic and asked me to paint it in oil for her.  I did.  It took me ages, painting at night after work, and foregoing weekend afternoon naps to get it done.

And, this is the rub:  Rebekah Joy Plett wrote, “When you buy something from an artist, you’re buying more than an object.  You are buying hundreds of hours of errors and experimentation.  You’re buying years of frustration and moments of pure joy.  You’re not buying just one thing, you are buying a piece of a hear, a piece of a soul.  A small piece of someone else’s life”.

not an artificial intelligence painting

If AI is a tool to be used by its owners for specific purposes and, as has happened, in about November 2022 artists sued an AI company for using copyrighted images to train their algorithms without asking for consent or offering compensation, the concept is no longer brilliant but an affront and breach of copyright and ethics.  Talk about flipping the bird to artists the world over.

It is as though the evolution of AI has spurred a flurry of all that is thoughtless and, quite frankly, despicable.  In this instance, I have to ask, “what are they thinking?” besides, quite obviously, of financial gain.  Has the “everyone gets a trophy” mentality extended to provide artificial talent, free of charge no less! to the less talented at the expense of artists who have spent years honing their skill?  Oi vey!  That is not progress but, rather, the pernicious descent of values.

Likewise, although less contemptuous, perhaps, are AI writing tools. For example, Copy.ai focuses on generating short-form marketing-related content in seconds.  It’s great for those with writer’s block and can save hours of research, brainstorming ideas, and help get the writing process started.  This particular app is free to use, but if you have a paid account, you have access to an accurate plagiarism feature, which will check and score your text within seconds.

Will AI writing tools replace writers?  Probably not, because at this stage AI doesn’t understand emotions and empathy and can’t check facts or create strategic content without human intervention.  AI might make some writers lazy, and easily replace poor writers, but if AI writing tools are used as a complement rather than a replacement, it will boost the works of the truly creative, good writers and simultaneously rid the world of inferior content.  A win-win situation if you ask me!

The fact remains, though, that AI writing tools cannot generate creative or original content. They work by using advanced natural language processing techniques and generate ‘new’ text by combining and rephrasing existing text.  Obviously, the existing text was originally written by someone and, following the lawsuit mentioned a few paragraphs ago, writers have begun to dispute intellectual property – another case of AI systems being trained on vast amounts of copyrighted work without consent, credit, or compensation.  Again, oi vey!

… and, lastly, the laughable

I have a profile on an international business and employment-focused social media platform.  I maintain my profile for a few reasons:  primarily, because it offers an online learning platform; secondly, it keeps me, well, linked in, I guess; and lastly, I keep an eye out for potentially fabulous jobs.

I enjoy the learning platform – it provides easy access to well-presented, informative, and educational courses and makes it easy for me to keep up with developments in my field.  The option to network is great, and reading posts is another convenient way to keep up with trends and opinions.  It’s the third option that I find doesn’t work very well…

Recruitment AI, designed to improve hiring practices by identifying more qualified candidates, seems to have a small glitch.  There is an option to set an alert for job advertisements that are well-suited to your profile.  I have set it.  The app then tags all suitable position and then it rates your suitability in comparison to the applicants that have already applied.

I have been placed in the top 10% for, amongst other equally unsuitable positions:

  • Manager, Citrus Farm;
  • Software Development Manager; and,
  • Patient Care Manager.

I know nothing about citrus farming, very little about software development, and I am certainly not registered with the nursing council, a prerequisite for the third position.  The first- and third-mentioned positions require that the candidate speak Xhosa.  I do not speak Xhosa.

This can mean only one of two things:  either the other candidates that applied are even less qualified than I am (scary thought), or the recruitment AI needs a little more work.

From a mind riddled with indecision about the advantages and disadvantages of AI, I leave you with this quote from Elon Musk: “I’m increasingly inclined to think there should be some regulatory oversight, maybe at the national and international level just to make sure that we don’t do something very foolish.” 

My thoughts exactly, chap!

Pin It on Pinterest