Dr Alok Aggarwal pioneered the idea of ‘data course of outsourcing (KPO)’, co-founded Evalueserve (4000+ workers), based IBM’s India Analysis Laboratory, and based Scry AI that builds proprietary AI merchandise for enterprises globally.
He has revealed 125 analysis articles, taught 2 programs on the Massachusetts Institute of Know-how (MIT), and has a Ph.D from Johns Hopkins College and a B.Tech from the Indian Institute of Know-how (IIT) Delhi.
On this dialog with Pankaj Agarwal, with insights drawn from his e-book The Fourth Industrial Revolution in 100 Years of AI from 1950 to 2050, Dr Alok Agarwal presents a compelling argument for why AI is not only one other technological pattern however a catalyst for a brand new industrial revolution. He delves into the historical past of commercial revolutions to grasp what makes AI stand out.
From steam engines to CPUs, every period has been marked by innovations that reworked societies. This episode provides an intensive evaluation of how AI compares to those previous improvements, whereas additionally cautioning towards the hype that surrounds it.
For entrepreneurs, the episode highlights the dangers of getting caught up within the AI hype with out growing strong mental property and suggests methods for creating high-value AI merchandise.
AI’s function within the Fourth Industrial Revolution
Dr Alok Aggarwal, with a outstanding profession at IBM, and because the founding father of E-ValueServe and Scry AI, has been on the forefront of AI’s progress.
“In each industrial revolution, a number of innovations come collectively to type new infrastructures,” Dr Aggarwal explains. “For instance, within the first revolution, steam engines fashioned the spine of change, whereas within the second, it was electrical energy. Right this moment, AI isn’t solely creating new infrastructure however can also be interacting with different applied sciences like IoT, blockchain, and gene modifying.”
AI stands aside from earlier industrial revolutions as a consequence of its capacity to deal with complicated duties that require reasoning and contextual understanding, making it a foundational know-how of our time.
Dr Aggarwal elaborates, “AI is already being utilized in greater than 2,500 operational use instances in the present day, and by 2050, I consider we’ll see it utilized in over 100,000 use instances.”
The hype and actuality of AI adoption
Pankaj, the host, factors out the thrill surrounding AI, significantly generative AI, and the way tech leaders like Sundar Pichai have in contrast AI’s influence to that of electrical energy. He asks Dr Aggarwal whether or not AI could have a extra pervasive influence than earlier technological shifts.
Dr Aggarwal responds with cautious optimism, “There may be a number of hype round AI in the present day, very like within the Nineteen Fifties and Nineteen Eighties. Nevertheless, it is very important keep in mind that all main innovations take time to combine into society. Simply because it took many years for electrical energy to succeed in half of the American inhabitants, AI adoption may even take time to mature totally.”
Whereas the media focuses closely on LLMs (giant language fashions) and GPTs (generative pre-trained transformers), Dr Aggarwal stresses that we’re nonetheless within the early phases of AI’s real-world functions.
“These fashions are glorious at producing textual content and fooling people with their proficiency, however we haven’t but solved the issue of the right way to totally combine them into production-level methods,” he remarks.
AI’s influence throughout industries
The dialog then shifts towards the sectors that can be most affected by AI within the coming years. Dr Aggarwal outlines that just about each business and division inside an organisation will expertise AI’s affect, from finance and procurement to advertising and marketing and gross sales.
“AI is outstanding at classification and sample recognition,” he explains. “In healthcare, for instance, AI can differentiate between whether or not an individual has pores and skin most cancers or not. In provide chain administration, AI is already detecting water leaks and optimising supply routes, whereas in banking, AI helps assess credit score dangers and enhance fraud detection.”
One significantly thrilling space that Dr Aggarwal highlights is clever doc processing, which may rework industries akin to insurance coverage and authorized companies.
“A lot knowledge is trapped in paper paperwork and PDFs. AI can course of and convert this unstructured knowledge into usable data, saving firms money and time,” he says.
Challenges and alternatives for AI entrepreneurs
When requested about recommendation for AI entrepreneurs, Dr Aggarwal gives sensible steerage on how startups can thrive within the AI house. He emphasises the significance of beginning with small, particular use instances and making certain the mixing of material experience alongside machine studying.
“AI is not only machine studying. Startups should perceive that AI requires each studying from knowledge and the appliance of skilled data. A giant purpose why many startups fail is that they neglect to include material experience into their methods.”
He additionally warns towards falling into the entice of over-relying on open-source fashions with out constructing proprietary mental property.
“You possibly can’t simply piece collectively open-source fashions and name it a product,” he cautions.
“To succeed, firms have to develop IP that provides them a aggressive edge, akin to enhancing accuracy by combining fashions or making use of domain-specific data.”
Trying forward: AI’s future potential
Because the dialog attracts to a detailed, Dr Aggarwal shares his perception that AI will finally play a extra vital function in society than electrical energy, although it can take time to understand this imaginative and prescient totally. He notes that India, with its huge pool of software program professionals and experience in knowledge annotation, is poised to turn out to be a significant participant within the world AI panorama.
In his parting ideas, Dr Aggarwal emphasises the long-term nature of AI innovation.
“It could take one other 10 to fifteen years for AI to turn out to be actually pervasive, however when it does, the world will see it embedded in each aspect of enterprise and every day life. This Fourth Industrial Revolution is in contrast to something we’ve seen earlier than,” he concludes.
Timestamps
03:55 – Historic evaluation of commercial revolutions
19:11 – The influence of AI on industries
34:05 – Navigating AI and mental property
45:27 – AI reworking companies in India