Last two and a half decades have seen phenomenal development in technology at an unprecedented pace. We can’t imagine a life without internet and phone today. Gen Z and Gen Alpha are amazed on how older generations survived without these.
At start of 21st century, Google
democratized internet and brought all the public knowledge available online, to
our disposal. Apple ensured every hand had a phone, thus getting this
information at our fingertips, literally. Facebook ushered in new ways of
connecting and interacting with people. Social media encouraged everyone to
broadcast photos, content and opinions. Digital revolution invaded our lives faster
than we could even realize.
These days, AI in general and Chat GPT in
particular is busy capturing our mind-space. It’s not just the awe of a superlative
technology, but also the fear of losing professional jobs that inspires and
worries people at the same time. Will AI cannibalize our jobs? What will we do
if machines do everything? Will machines become our new masters? Are the days
numbered for human art? Are we missing the bus? Our worries and queries have no
end.
To get our answers, it is essential we
understand what AI actually is. In simple terms, AI is a collective reference
to computing algorithms which learn patterns from historic data, and predict
the outcome of a future (unseen) data by matching it to a pattern it has
already learnt. GPT-3, which belongs to the “Generative AI” family and powers
ChatGPT, is an AI model that is said to be trained on the entire information
that was available on internet till June 2021. It is a language model, meaning
it is trained on predicting the next word, e.g. for “Twinkle twinkle little “,
the next most probable word (out of thousands of other English words) is “star”
and that’s what it will predict based on its learning.
The current AI makes machines smart enough
to perform mundane, laborious and repetitive jobs, or at least that is what
they are being considered for at this moment. Digital vulnerabilities like biased
and influenced results, limited abilities to filter out fake content, knowledge
limited to training data source, regulatory issues, etc. still keep many away
from embracing AI completely.
So, do we really need to worry about AI
eating our jobs?
In his book “The Atomic Habits”, James
Clear mentions that humans overestimate the situations in near future where
chances of occurrence are minimal (like the flight you travel in being
highjacked) and underestimate the vulnerabilities that are far ahead in time,
but sure to occur (like obesity due to unhealthy lifestyle).
Fear of losing jobs due to AI is like
fearing of being hijacked. In the last 300 years, every major revolution has
seen such discussions soaring in, machines cutting hands in Industrial
Revolution, Computers killing jobs in Digital revolution, etc. There was a hue
and cry about Amazon, Walmart and the likes cannibalizing the General Stores we
find in every corner of street. However, the fact remains that shops, both
online and in brick and mortar, can coexist and flourish together. Invasion of
Robotics in manufacturing began 3 decades ago, however there is no record of
major job cuts in the sector globally.
With AI, the mundane repetitive jobs are
sure to get transferred from humans to machines. At the same time, humungous
data generated every second allows opportunities for more data exploration and
analysis, thus increasing the breadth of its application, hence more jobs. Thoughtful
application of AI will only relieve humans from the monotony of unproductive
jobs and allow time to focus on more interesting and productive things.
Computer code generated by “Co-Pilot”
(Gen-AI that generates application code) is too generic to fit directly in a
heavily customized solution. We will still need software engineers to translate
the generic to specific. Again, Gen AI serves as the Co-Pilot (NOT an
auto-pilot) to the software developer, thus allowing the latter to focus more
on logic than syntax.
AI models use Neural Networks to learn
patterns and encode this information in the form of numbers. Capabilities are
limited to the corpus that they are trained on, hence anything beyond that is
out-of-scope. They have knowledge, but lack wisdom. That said, such models are
found to be less effective on scenarios that deal with unknowns.
Consider situations like Covid. Everyone
was dealing with “unknown” unknowns. Neither was any data available to train
models, nor did the “pattern” match with anything already learnt. Dealing such
unknowns requires human judgement developed out of varied experiences,
collective knowledge, emotions, empathy, social awareness, study of cultural
influences and social behaviour, local laws, hunches and a lot more factors.
These may not translate entirely into measurable dimensions that a model can learn
from. All such work will always stay human, where AI can be an obedient and
efficient assistant.
In near future, machines might automate
creation of medical transcripts by capturing details from conversations between
a doctor and patient. It might create marketing content with impressive narration
and images. It has already started writing articles and creating professional
presentations. It has also begun answering customers’ calls, chat messages and
playing the translator. It’s driving cars. It suggests machine failures and
also fuels our OTT indulgence by suggesting content.
All such jobs are sure to get rid of
humans. But ideation, anything that requires empathy (consulting, counselling,
therapy), human judgement (disease diagnosis, legal judgements, policy making)
will and should always remain human. Emotional quotient is proven to be a major
determinant in professional success, which, thankfully hasn’t yet been drummed
into the machines. We will need humans to motivate us, not machines.
AI is like electricity; it has potential
that can only be harnessed through smart application. Applying AI to where it
adds value will also remain a human thing.
I feel certain things should consciously remain
human. Art should emerge from human mind. Fun of creation should stay human. We
shouldn’t let AI rob from us our choices, ability to think and thrill of
discovery.
While embracing AI, we should also be concerned about the latter part explained by James Clear – “Humans underestimate the vulnerabilities that are far ahead in time, but sure to occur”. AI can be a great servant but a terrible master. Controlled and ethical use of AI can make it a great servant, while its overuse may handicap us forever.
With every major revolution emerges an inherent evolution that is relatively slow, silent, but which creates an impact over time. Our conscious choices can steer this evolution towards a better world.