Whether watching anime, reading comics, enjoying science fiction movies, or attending a magic show, each experience contains an element that captivates. A camera flash, a puff of smoke—the illusion of something impossible made real. That’s the old-school magic. 

But today, one only has to type a prompt and watch an AI model produce poetry, presentations, and even predictions. AIM decided to ask leading developers and tech professionals if they thought there was anything magical about AI. The point was to see how people perceive this breakthrough technology. Some called AI surprising, some called it powerful, but almost all stopped short of calling it “magic”.

Science, Not Sorcery

At the Great International Developer Summit, Stephen Chin, VP of developer relations at Neo4j, told AIM, “No. AI is science, and it’s programming, software development. The magic is that sometimes we’re surprised at the quality and the results of what comes out of models.” 

He acknowledged that results can be unexpectedly impressive, but that simply reflects the evolving maturity of tools and models. According to Chin, AI’s actual value lies in enabling developers to be “more efficient” and helping them build “better and more innovative applications.”

A similar sentiment came from Brian Lanigan, SVP of global partner ecosystem and emerging product sales at SentinelOne, who told AIM, “Is AI magic? No, in my mind, AI cannot be magic.”

In cybersecurity, where stakes are high and threats evolve rapidly, Lanigan sees AI as an essential augmentation, one that’s indispensable, but still grounded. “We, as cybersecurity professionals in the industry, have to use AI to fight AI,” he said.

Dean Teffer, VP of AI at Arctic Wolf, offered a slightly more nuanced take. “Yeah, I don’t believe in magic, so I don’t think it’s magic,” he said. 

Still, he confessed to being “shocked” by GPT-3’s capabilities and admitted even researchers from the earlier era hadn’t expected the scale of emergent behaviour that followed. “We’ve all had gee-whiz experiences with generative AI,” Teffer noted. But with time, he said, one can trace back the mechanisms. “You can sort of work backwards and figure out” how models work.

Magic, From a Certain Point of View

Daniel Oh, senior principal developer advocate at Red Hat, didn’t give a straight yes or no. “Is AI magic? I don’t know,” he told AIM, before launching into an anecdote about a friend who now uses ChatGPT to write polite emails in English, a task he once struggled with. 

For Oh, that feeling of enchantment is less about mystery and more about seamless integration into daily life. “Maybe I’m going to say that is AI magic for non-technical people.”

Jeff Green, chief development officer at Arctic Wolf, agreed that AI’s strength lies in making complex tasks easier. He pointed to its ability to analyse vast datasets and detect cybersecurity threats with a speed and precision that would stump most humans. “That, for a human, is hard to do, very hard to do,” he said, calling the technology “super impressive”, if not magical.

Recently, we’ve also seen people increasingly rely on ChatGPT as a personal health assistant, sometimes with life-changing outcomes. From helping users diagnose chronic issues like back pain and jaw problems to offering dietary suggestions during electrolyte imbalance, ChatGPT is being credited for providing clarity when traditional medical advice falls short. Some users even claim it helped them avoid serious emergencies or better understand rare diseases.

OpenAI leaders like Greg Brockman and Sam Altman have acknowledged this growing trend, with Altman expressing hope for broader healthcare access through AI. 

Furthermore, AI-driven sports analysis in India is gaining momentum, aiding performance, injury prevention, and talent scouting, especially in top-tier leagues. Startups like ScoutEdge aim to bridge data gaps in tier 2 and 3 regions to take this further.

This is just a few examples of how AI has been gradually integrating with daily life activities. So, is AI magic? Not quite. 

For instance, AI surveillance is increasingly used in India for crowd management at religious and public events, using facial recognition, behavioural anomaly detection, and drones to predict and prevent disasters. 

Startups like Innefu Labs and IG Drones provide real-time analytics through multimodal AI systems. Despite successful deployments at festivals like Ratha Yatra in Ahmedabad and Puril, failures persist—fatalities still occur due to stampedes. 

There are many fundamental challenges yet to be addressed to make AI as effective as users hope it to be.

But it may feel like magic, especially when it works so well, so fast, and so often without one fully understanding why. Perhaps that’s the modern illusion, not deception, but the unexpected element crafted from code.

The post Is AI Magic? appeared first on Analytics India Magazine.