
Sign up to save your podcasts
Or


Just two years ago, the launch of GPT-4 required the support of hundreds of experts, intensive computing power, and vast coordination. Today, that same process could be executed by a team of five. That is not an exaggeration. It is a signal of just how rapidly AI development is accelerating. In this episode of Tech Talks Daily, I'm joined by Gagan Tandon, Managing Director of AI & Data Services at TELUS Digital Solutions, to explore what is behind this extraordinary shift.
From the evolution of AI chip design to the efficiency gains from smarter algorithms and open-source tools, Gagan offers a clear and grounded look at how AI training has become more accessible and less resource-intensive. But the bigger question is, what can we do with that efficiency? For Gagan, the answer lies in impact. Whether it is addressing bottlenecks in global supply chains, enhancing diagnostics in healthcare, or reducing fraud in financial services, these advances in AI development are not just theoretical. They are opening new doors for innovation across industries.
We also explored how TELUS Digital Solutions supports enterprise transformation through its Fuel iX platform, which helps organisations transition from GenAI pilots to scalable, secure, ethical, and practical production systems. Gagan shared examples of how large-scale transformation can be approached with agility, especially when AI is aligned to specific business outcomes rather than simply layered on top of existing systems.
This conversation goes far beyond model training. It is about what's next in AI operationalization, what business leaders should prioritize when it comes to adoption, and how to avoid common pitfalls as technology continues to evolve faster than many are prepared for.
So how did we go from 300 engineers to five? What should enterprises be doing right now to take advantage of this momentum? And how can we make sure that efficiency does not come at the cost of responsibility? Join us as we explore the answers.
By Neil C. Hughes5
198198 ratings
Just two years ago, the launch of GPT-4 required the support of hundreds of experts, intensive computing power, and vast coordination. Today, that same process could be executed by a team of five. That is not an exaggeration. It is a signal of just how rapidly AI development is accelerating. In this episode of Tech Talks Daily, I'm joined by Gagan Tandon, Managing Director of AI & Data Services at TELUS Digital Solutions, to explore what is behind this extraordinary shift.
From the evolution of AI chip design to the efficiency gains from smarter algorithms and open-source tools, Gagan offers a clear and grounded look at how AI training has become more accessible and less resource-intensive. But the bigger question is, what can we do with that efficiency? For Gagan, the answer lies in impact. Whether it is addressing bottlenecks in global supply chains, enhancing diagnostics in healthcare, or reducing fraud in financial services, these advances in AI development are not just theoretical. They are opening new doors for innovation across industries.
We also explored how TELUS Digital Solutions supports enterprise transformation through its Fuel iX platform, which helps organisations transition from GenAI pilots to scalable, secure, ethical, and practical production systems. Gagan shared examples of how large-scale transformation can be approached with agility, especially when AI is aligned to specific business outcomes rather than simply layered on top of existing systems.
This conversation goes far beyond model training. It is about what's next in AI operationalization, what business leaders should prioritize when it comes to adoption, and how to avoid common pitfalls as technology continues to evolve faster than many are prepared for.
So how did we go from 300 engineers to five? What should enterprises be doing right now to take advantage of this momentum? And how can we make sure that efficiency does not come at the cost of responsibility? Join us as we explore the answers.

1,287 Listeners

537 Listeners

1,640 Listeners

1,090 Listeners

164 Listeners

111 Listeners

303 Listeners

334 Listeners

269 Listeners

207 Listeners

9,936 Listeners

5,509 Listeners

349 Listeners

93 Listeners

608 Listeners

0 Listeners

0 Listeners

0 Listeners

0 Listeners

0 Listeners

0 Listeners

0 Listeners