Our AI Journey
Artificial Intelligence (AI) has taken over most business conversations. The progress is astounding. The use cases are promising. The possibilities are intriguing. The concept is not new. We’ve been toying around with AI since the 1950’s. Research and experimentation slowed down a bit in the early 2000’s and now it is skyrocketing for obvious reasons. The capabilities of data capturing, storing, and computing are not a limitation anymore. Let’s skip the rest of the history lesson and focus on today. AI is here. It is happening. And it is leaving mixed feelings around its presence, its essence, and its role. What do we do with it? How do we use it? How much can we trust it? How can it help our businesses? And when will it take our jobs?
The Exciting Part of AI
Seeing images and sceneries get constructed in front of our eyes and resulting in flawless visuals is breath taking. Receiving on-point responses to our various inquiries and engaging in a constructive conversation with a web application is fascinating. Viewing personalities recorded on video having a discussion they never had in real life is jaw dropping. This is where exciting starts. But it is not what AI can only do. It can do real work making us more productive. It can make supply chain decisions. It can make medical diagnosis. It can make personalized curriculums for students based on their individual capabilities.
At the core of it, AI is possible because of two things: an advanced algorithm and data—lots of data. AI algorithms differ from what we’ve traditionally been accustomed to in previous technology innovations. They do not rely on if/then/else logic. There is no human-inferred logic at all. It is left to the machine to draw decisions based on its deep learning with the help of neural networks and data. AI feeds on data. The more the data, the better AI is in developing synopsis and characterization.
AI has become exceptionally good in making decisions on content development, customer insight, production processes, and even medical diagnosis. It is one thing for AI to do what we do. It is another thing for AI to do it better than us. AI now has lower error rates than those of a human. That’s where excitement turns into fear.
The Frightening Part of AI
The fear from AI falls into three camps. The first is around the bias generated by AI. AI reflects on what it learns from data. It is hard to ensure that data is inclusive of all situations. If it is a loan approval AI, is it influenced by demographics or environmental conditions? There is no standard validation of AI data. Thus, there is a concern that although AI can be precise based on the data feeding it, it can be biased. And if there is probability of bias, how can we trust its objective reasoning when it comes to sensitive areas?
The second fear from AI is loss of jobs. If AI is destined to be as good and better than a human in resolving issues, accomplishing tasks, and making decisions, it is only time before AI replaces people in the workforce. Forecasts range from 10% to 70% job loss by 2030. We haven’t been good at forecasting the impact of previous innovations; nonetheless, there will be change. It is not necessarily all bad. There will be opportunities with roles never existed before, like prompt engineering. Humans can adapt to changes. The only difference here is that it is happening at an astounding speed.
Living with AI makes the third fear. AI as we know it today does not yet fit its theoretical definition of mimicking how a human brain works in thinking, rationalizing, and directing behavior. People can do their taxes, paint a canvas, engage in jury duty, and hold a conversation on ancient literature. We may not do all of them with expertise, but we can cross-learn and think about a multitude of topics. In contrast, AI today is deep and narrow. AI image generator cannot review financial statements. But then again, it is unlikely we have a judge who is also a surgeon. We don’t know if, or when, AI capabilities will eventually converge with that of humans. But if they do, what will become of our purpose? Even with the current state of multiple specialized AI solutions where each can generate efficiencies in respective areas, will a business with a single owner employing AI to serve business operations be a state we are willing to accept and live with?
The Curiosity to Keep AI Going
We’ve been curious about AI since 2011 when IBM Watson beat Jeopardy champion and earlier in 1997 when IBM Deep Blue supercomputer beat the world’s champion in chess. The fears may be big, but the curiosity is bigger. What else is possible? And how much is possible?
Neither of the fears above has a strong rebuttal, so they are all valid. AI can redefine who we are and what we do, especially when AI starts generating and managing new AI without any human guidance. Time magazine’s End of Humanity cover and an AI Pause request letter from AI enthusiasts make these fears more real. But it is all hypothetical. We don’t know what we don’t know. That’s why it is hard to pause now. We are all curious to know. Curiosity will continue to drive us to learn, experiment, and innovate. The result can be something we never dreamed of before. Will AI help us travel in time? Will we gain the knowledge to explore outer galaxies at the speed of light because of the extra artificial neural power we will have? Will this become our new purpose and identity rather than a job title?
What to Do Today
As an individual, it is important to get comfortable with AI. If not to use, then to know well. We will all be touched by it in the near future, if not already, one way or another. We should ask questions—not out of fear but out of curiosity. And we should actively participate in developing it and making it better as it pertains to our domain and our expertise.
A business or an organization should engage in AI adoption to advance operations efficiency and decisioning precision. But in parallel, it should equally focus on social responsibility. Social responsibility is a mandatory track for the diffusion and evolution of AI. Understanding anticipated impacts and actively working to transition societies to a workable future state is the responsibility of every business benefiting from AI. The work around social responsibility is lagging, hence the fear. For it to catch up, AI advancement does not have to slow down or pause. We just have to turbo charge social awareness and practical solutioning. That goes beyond setting limiting laws and protocols. Ironically, AI may be an option to help set its own social guardrail.
In practice, an organization needs to find its AI opportunities. After understanding that AI is not a blanket solution across the chain of operations, targeting efficiencies in a particular area (acquisition, production, fulfillment) is where progress and success can be achieved. From there, a respective model or algorithm is identified and secured. The marketplace has sufficient models to choose from with simple setup. Third party data can be leveraged for the AI, but if an organization wants or depends on its distinctive customer base insights, then increased efforts of first party data collection, curation, cleansing, storing, and analyzing are necessary. This will empower their AI solution to provide tailored decisions and directions supporting their unique efficiency and growth trajectory.