Undated
Darko Todorovic is the Chief Technology Officer of HTEC Group.
“Today's guest is Darko Todorovic, Chief Technology Officer at HTEC Group”
Topic
Darko Todorovic
15 facts from 1 episode, in date order. These are what speakers said on the shows, each linked to the moment it was said. They aren't independently verified.
Undated
Darko Todorovic says organizations typically form an AI task force inside the CIO or CTO office to find processes to augment with AI.
“you have a task force that is formed inside of a C I O office or a C T O office.”
Undated
Darko Todorovic says organizations often adopt AI after seeing other companies report efficiency gains of 200% or 400%.
“they see that that organization implemented A I and they have efficiency of 200% or 400%, et cetera”
Undated
Darko Todorovic says proofs of concept succeed partly because they are staffed with the organization's most senior people, who have strong judgment.
“You chose the right process, you chose the most senior people in the organization, the best people that you have, that have already a decision making framework, good judgment.”
Undated
Darko Todorovic says digital transformation took about ten years, while the AI shift is condensed into roughly eighteen months.
“We we've seen that process, but that process was ten years long. Now everything is condensed in eighteen months.”
Undated
Darko Todorovic says AI investment should be assessed as an upfront non-recurring engineering (capex) cost plus a per-unit execution cost.
“One component is initial N R E investment, so basically capex, and then you have to see how much is execution of these agents or A I tools is going to cost you per each unit of measure.”
Undated
Darko Todorovic says a well-architected AI solution lets organizations monitor token consumption per agent and per person and see the monthly bill.
“you can monitor the token consumptions per agent per person. You can actually see what is the the bill at the end of the month”
Undated
Darko Todorovic says a sophisticated team can deliver the same AI solution while spending fewer tokens than a less skilled team.
“sophisticated team is going to give you the same solution for the less amount of spend, or less, uh, uh, smaller amount of tokens spent, which is really important.”
Undated
Darko Todorovic expects AI costs to fall over time and says serious adopters will not worry about cost in the first couple of years.
“serious organizations that see serious advantages of using A I are not going to be worried about the cost in the first couple of years, because the cost is going to go down over time.”
Undated
Darko Todorovic says ChatGPT's answers have improved markedly in comprehension compared with two years earlier.
“if you look at what Chat G P T gave you as an answer two years ago compared to what you're getting now, it's like really, really different comprehension”
Undated
Darko Todorovic says a successful AI agent in one department can lead other departments to demand it, potentially multiplying payments to foundation model companies tenfold.
“I'm giving to some of the foundational model companies, uh, x amount of money, and now I have to to give ten x.”
Undated
Darko Todorovic says AI readiness should be assessed on two dimensions, the first being technological and data readiness and the second being organizational readiness.
“One dimension is uh technological readiness and data readiness.”
Undated
Darko Todorovic says standard enterprise software such as HR, procurement and ERP systems made different organizations look alike because customization was costly.
“all of the standard processes like H R process, procurement processes, E R Ps, et cetera, were built in a way that when you implement them, every organization is looking like the other organization.”
Undated
Darko Todorovic attributes the growth of software-as-a-service businesses in the last decade to how hard it was to customize traditional enterprise software.
“That's why you had all of these S A S businesses exploding in the last decade.”
Undated
Darko Todorovic says a knowledge management layer on top of company data is important so multi-agent systems can share the same context.
“Knowledge management is really important if you want to have multi-agent systems that can share the same context”
Episodes