Great day at Citi’s Gen AI Summit. Very insightful and factual panels, with takeaways consistent with customer conversations we’ve had in the last quarter: * Most customers are still very much in the Gen AI PoC stage today, and continue to be cautious about deploying anything at-scale in production * Many of those in production are primarily doing chatbots focused on internal facing application and customer service; feels very much in the early stage of the first innings * For those in production, the most time is spent in the last 10% of getting a model into production - many last minute wrinkles to iron out * Counterintuitively, regulated industries are moving faster on Gen AI use cases, because the in-house data is much cleaner (thanks to adherence to regulation) * If you thought your cloud bill was high.. wait till you see your AI bills - will force enterprises to think doubly hard about use cases Good seeing Prayank Swaroop Burak Arik and many others there! Big kudos in pulling this together Matt Carbonara ✅ Vibhor Rastogi Çağla Kaymaz!
Great summary and takeaways Aman. You have touched upon some important trends that are impacting the successful adoption, operationalization and financial justification of AI initiatives across firms. The quote ""Hype is the bane of all technology; it oversells promise and under-delivers results" summarizes it all.
Cogent analysis as always, Aman. The last bullet in particular. Free ain’t free. And with Anthropic saying their Claude 5x model may cost as much as $10b to train with $100b models in future enterprise buyers increasingly sensitive to future costs, slowing uptake.
This is what gen ai investors and proponents miss..Hype to implementation and cost of ai.
Simon Wu
3w
If AI bill and (ultimately) ROI bar is that much harder - does that make adoption that much harder? Seems like we need a clear map on processes in order to make rational decisions!
Çağla Kaymaz
3w
Great to see you Aman, thanks for joining us!
James Probst
3w
Very insightful, Aman Rangrass. It is not surprising that chatbots are the first frontier. They make logical sense and have a very clean use case and user base that would benefit. Great insight about regulated industries moving faster.
Sr. Client Engagement Manager - QA & DevOps Practice | AI Quality Engineering | Quality Assurance Excellence | Innovation in Quality Assurance | Business Impact Through Testing
3wAbsolutely fascinating insights from the Gen AI Summit at Citi! It’s intriguing to see many customers still in the PoC stage, cautiously approaching at-scale production. It’s fantastic that regulated industries are accelerating Gen AI use cases thanks to cleaner in-house data. The impact of AI regulations forcing enterprises to scrutinize use cases is definitely a critical point. Big congrats to Matt Carbonara, Vibhor Rastogi, and Çağla Kaymaz for organizing such a stellar event!