Inside AI on Tap London: Episode 3
From AI adoption at HSBC to model costs and choices, a discussion of practical constraints and why the people doing the work must shape process change.
Photos from the event · 1 of 5
On 8 September, Prescient Labs sponsored the third London AI on Tap gathering alongside We Are Agentic. The panel brought together people at different stages of putting AI to work, which made for a more useful conversation than another discussion of what AI might eventually do.
There was a panel of interesting speakers across the AI consultancy world, as well as real-life and client experiences. This included Ben Iceton at HSBC, who spoke about the start of its AI implementation journey. In a regulated environment, adoption has to work within real controls and responsibilities. His perspective brought the discussion back to the practical conditions in which a promising tool has to operate.
The open discussion moved from adoption to the technology itself. People considered open and closed-source models, GPU costs and how to weigh a narrowing performance gap against the cost and constraints of running a model. There was no single answer that would fit every organisation. The choice becomes clearer when the intended use, economics and operating environment are specified.
Our CTO, Sarah, closed the panel with a related point about how change begins. A leadership team may know what it wants to improve, but the people doing the work know why a process has developed its particular steps and exceptions. Bringing those perspectives together early makes it possible to question the right things before changing them. Otherwise, a redesign can remove a step that looked redundant from a distance but served a real purpose.
Thank you to AI on Tap, the speakers and everyone who took part. We cannot wait for the next London gathering on 13 October. For more AI on Tap events in Brighton, Sydney, Berlin and Manchester, see the AI on Tap website.