From chatbots whispering sweet nothings to predictive algorithms knowing you better than your therapist, AI is the shiny addictive toy everyone wants to play with, whether they understand it or not.
The allure of AI is akin to that of religion in its heydays; AI has captured the public imagination. Need to sort your emails, plan your retirement or write a love letter? AI to the rescue. It is the digital messiah, offering convenience and efficiency wrapped in an easy-to-use user interface.
Studies show that global AI adoption in businesses shot up by 270% between 2015 and 2022, with the AI industry projected to hit $1.8 trillion by 2030. On the social side, from viral AI-generated creations to ChatGPT penning bedtime stories for kids, AI is a pop-culture movement as much as a tech revolution.
The numbers are miraculous. ChatGPT reached 100 million users in two months, while it took Instagram 2.5 years to do the same. Even TikTok, the last big smash hit, needed nine months to get similar numbers. Threads did it in just over 4 days, but is yet to develop a cult-following like ChatGPT.
Granted, AI offers significant uses. Using AI-powered data analysis, Google Flood Hub provides free flood alerts up to 7 days in advance in 80 countries. The Ocean Cleanup organization employs AI to create detailed maps of ocean litter in remote locations, raising plastic-removal efficiency.
But AI cheerleaders get carried away. Poverty, inequality, world hunger, relationships… the list goes on. Granted that technology can assist problem-solving, but portraying AI as a miracle worker is a disservice to AI.
Let’s shine light on a few epic AI failures. Tesla’s Autopilot has been linked to 13 fatal accidents in 2024 alone; hundreds of accidents and over 40 fatalities since it became street-legal.
IBM Watson for oncology failed to improve cancer care. The project cost $62 million for M.D. Anderson Cancer Center without achieving its goals. Chatbot Grok falsely accused NBA star Klay Thompson of vandalism in April 2024. One could go on.
Corporations are not laughing their way to the bank with their AI investments either. A Gartner report indicates that about 30% of Generative AI projects will be abandoned after the proof-of-concept stage by the end of 2025.
Companies struggle to demonstrate value from these projects, often requiring huge upfront investments. A RAND Corporation report reveals that over 80% of AI projects fail, double the failure rate of traditional IT projects. Common causes include miscommunication of goals, inadequate data prep and lack of necessary infrastructure.
For a long time, I have believed that the ills of AI lie in its very name. ‘Artificial intelligence’ has an implicit bias. It implies the possibility that machines will develop some form of consciousness and emotions, acquire a human-like ‘personality,’ and ultimately overcome human limitations—and voila, AGI!
To some experts, an easy way to fix AI is to rectify the name itself. Call it by what it actually does: Systematic Approaches to Learning Algorithms and Machine Inferences, or Salami, in short.
That changes perceptions and expectations. Can Salami develop any consciousness? What about emotions and a personality? Will Salami replace a human? Can you fall in love with a Salami? Maybe this could show us how ridiculous expectations of AI are.
No conversation on AI is complete without talking about GenAI, the tech industry’s recreational weed. Its great accomplishment was democratizing the use of AI quickly. And therein lies the issue.
A recent Blueoptima study tackled the question of GenAI’s impact on developer performance by analysing 218,000+ developers across 880 million+ ‘commits’ (with control groups and all).
The results? A modest but consistent 4% productivity boost without sacrificing code quality, which would reduce a team of 100 to 96 engineers (or boost output by 4%) Good, but not game-changing. It also found that only 1% of developers committed GenAI code without “significant rework.”
The study concludes that while GenAI tools can enhance software developer productivity, the impact depends on how they work with software development workflows. It’s most effective when AI augments rather than replaces human expertise.
The latest DORA Report found AI tools marginally upped individual productivity but reduced overall team productivity. Can organizations not get the same 4% with better planning, project management and good old common sense? Not to mention reducing CO2 emissions. Just something to think about.
AI, like religion, is what we make of it. Let’s create something that does not require digital exorcisms in the future. Let’s find a balance. Let’s approach AI with wisdom to recognize its benefits and scepticism to question its limitations. Think of it as being spiritually aware but not mindlessly faithful.
The author is a technology advisor and podcast host.
