Enterprise AI, 2023 to 2024
From Pilots to Copilots
scaling enterprise generative AI with data, controls and measurable value
Fractional Head of AI, Ecaveo
Abstract
Three days ago an enterprise product with contractual data terms became available, which changes the procurement conversation and leaves the evidence where it was. Most large organisations now hold a collection of pilots that were never designed to be compared with one another, and a growing expectation from their boards that something should be scaled. This paper sets out how to convert that collection into a governed portfolio. It proposes five stage gates, a reference architecture for an assisted workflow, a balanced value scorecard and an explicit scale, hold or stop decision with written thresholds. The design principle throughout is that scale is a portfolio decision supported by evidence, and a compelling demonstration is not evidence. A worker-centred section sits in the middle, where it can affect the design. The available international evidence on artificial intelligence in workplaces reports both improved job quality and material concern, from a survey conducted before these systems existed in their current form. Used carefully, it supports an argument about consultation and task redesign. Used as evidence about copilots, it supports nothing at all.
Read the paper
This is the reading edition, complete and unabridged. The designed edition, with the full illustrated layout, is available from the original publication page.
Cite this paper
Download RIS Download BibTeX RIS opens in Zotero, Mendeley, EndNote and RefWorks
BibTeX, inline
@techreport{forrest2023frompilotsto,
author = {Forrest, Paul},
title = {From Pilots to Copilots: scaling enterprise generative AI with data, controls and measurable value},
institution = {Ecaveo Services Ltd},
series = {Enterprise AI, 2023 to 2024},
year = {2023},
month = {08},
pages = {21},
url = {https://papers.ecaveo.com/from-pilots-to-copilots-scaling-enterprise-generative-ai/}
}