# Why isn't AI adoption faster?

Source: https://www.youtube.com/watch?v=LToW1rQ6rVA
Recap page: https://rapidrecap.app/video/LToW1rQ6rVA
Generated: 2025-11-09T12:32:51.741+00:00

---
## Quick Overview

AI adoption lags primarily because organizations are stuck in a "pilot purgatory" where they successfully automate low-value, human-centric tasks, leading to insufficient demonstrable ROI and a failure to address the true bottleneck: integrating AI into core, high-leverage business processes. The biggest disconnect is the focus on easily quantifiable metrics like time saved per task (e.g., 80 hours saved per week on a task that only takes 8 hours), rather than measuring the actual impact on bottom-line KPIs or overall system throughput, which stalls executive buy-in and prevents scaling beyond pilot projects.

**Key Points:**
- The primary barrier to faster AI adoption is organizations getting stuck in "pilot purgatory" (0:21, 6:06), focusing on small, easily quantifiable wins rather than systemic integration.
- A major disconnect is measuring time saved on low-leverage tasks (e.g., saving 80 hours on an 8-hour task) instead of real ROI or overall system throughput (4:59, 7:59).
- The Gartner report mentioned indicates that 30% of generative AI projects will be abandoned due to unclear business value (5:58).
- The key metric cited for success should be total throughput/system efficiency improvement, not just individual task time savings (8:15, 17:00).
- The speaker identifies two main deployment types: Horizontal AI (like Copilot/ChatGPT) and Vertical AI (highly targeted, high-leverage interventions) (7:08, 8:50).
- Legal contract review time reduction from 12 weeks to 10 minutes is cited as an example of a high-leverage win that should be prioritized (11:41).
- Leaders (CIOs, CFOs) often focus on compliance, risk, and governance (2:38, 13:12), rather than the transformative, high-leverage wins AI can provide.

![Screenshot at 0:05: The video title card visually sets the theme: "Why AI Adoption Lags," featuring a circuit board brain graphic and a prominent caution tape warning sign.](https://ss.rapidrecap.app/screens/LToW1rQ6rVA/00-00-05.png)

**Context:** The speaker, David Schapiro, discusses why the adoption rate of Artificial Intelligence (AI) across various industries and nations is slower than expected, drawing on insights from recent research, including a Gartner report. The context centers on the gap between the immense potential of AI technologies, like generative AI and Copilot tools, and the actual, tangible value being realized by enterprises, particularly small and medium-sized businesses.

## Detailed Analysis

The core issue hindering faster AI adoption is identified as "pilot purgatory," where companies successfully implement AI for small, easily measurable tasks but fail to integrate it into core, high-leverage processes, thus failing to achieve meaningful ROI. The speaker points out that many metrics used to justify AI investment, such as saving 80 hours on a task that only takes 8 hours, are vanity metrics that don't reflect true business value or overall system throughput. Citing a Gartner report, the speaker notes that 30% of generative AI projects are abandoned due to unclear business value (5:58). The speaker contrasts two deployment paths: Horizontal AI (like general tools such as Copilot) and Vertical AI (highly targeted, high-leverage interventions). True value comes from the vertical approach, like drastically cutting legal contract review time from 12 weeks to 10 minutes (11:41). Furthermore, leaders like CIOs and CFOs often focus on immediate compliance and governance risks, neglecting the transformative potential. The speaker concludes by advocating for measuring true system efficiency gains rather than just local time savings, emphasizing that effective AI implementation requires focusing on bottleneck removal, not just automating existing human-centric tasks.

### AI Adoption Context

- AI adoption lags due to organizations getting stuck in pilot purgatory
- Focusing on easily measurable, low-value tasks
- Failing to achieve meaningful ROI
- Citing Gartner data that 30% of generative AI projects are abandoned (5:58).

### The Two Deployment Paths

- Horizontal AI involves deploying general tools like Copilot to everyone (7:11)
- Vertical AI involves highly targeted, high-leverage interventions that address core bottlenecks (8:55).

### Flawed Measurement

- The primary metric cited is often time saved on tasks, which can be misleading (e.g., saving 80 hours on an 8-hour task) (7:55)
- True measure should be total system throughput improvement (8:15).

### Leadership Focus

- CIOs/CFOs often focus on governance, risk, and compliance (13:12)
- They fail to see the immense, transformative value of AI (15:15)
- The focus should shift from just keeping busy to demonstrably increasing efficiency.

### The Biggest Bottleneck

- Legal contract review time reduction is a clear, high-leverage win (11:41)
- The bottleneck shifts from coding to code review/deployment (19:05)
- The real constraint is realizing value, not just automating tasks.

### Conclusion & Call to Action

- Organizations must be human-centric, empowering humans rather than simply automating existing processes (21:27)
- Leaders must prove value through concrete metrics, not just busywork automation (23:03).

![Screenshot at 0:05: The video title card visually sets the theme: "Why AI Adoption Lags," featuring a circuit board brain graphic and a prominent caution tape warning sign.](https://ss.rapidrecap.app/screens/LToW1rQ6rVA/00-00-05.png)
![Screenshot at 0:38: Speaker David Schapiro begins outlining the context, noting he consults with leaders across various industries and capacities about AI adoption challenges.](https://ss.rapidrecap.app/screens/LToW1rQ6rVA/00-00-38.png)
![Screenshot at 1:11: The speaker introduces the concept of Horizontal AI \(like Copilot\) versus what businesses should be aiming for regarding measurable value.](https://ss.rapidrecap.app/screens/LToW1rQ6rVA/00-01-11.png)
![Screenshot at 2:23: The speaker identifies the need to unpack the eight points of research, emphasizing that the core issue is the gap between expectation and reality.](https://ss.rapidrecap.app/screens/LToW1rQ6rVA/00-02-23.png)
![Screenshot at 3:54: The speaker summarizes the research into eight distilled points, indicating the structure for the rest of the presentation.](https://ss.rapidrecap.app/screens/LToW1rQ6rVA/00-03-54.png)
![Screenshot at 5:07: The speaker identifies the biggest disconnect: measuring time saved on low-leverage tasks instead of focusing on throughput or core business impact.](https://ss.rapidrecap.app/screens/LToW1rQ6rVA/00-05-07.png)
![Screenshot at 11:44: A slide or graphic likely appears referencing the JP Morgan COIN project, used as an example of high-leverage automation that drastically reduced contract review time \(12 weeks to 10 minutes\).](https://ss.rapidrecap.app/screens/LToW1rQ6rVA/00-11-44.png)
![Screenshot at 13:13: The speaker discusses the negative focus leaders often place on governance, risk, and compliance \(GRC\) when evaluating AI projects.](https://ss.rapidrecap.app/screens/LToW1rQ6rVA/00-13-13.png)
![Screenshot at 15:54: The speaker highlights that many companies focus on time saved as a vanity metric, contrasting it with the actual impact on the bottom line.](https://ss.rapidrecap.app/screens/LToW1rQ6rVA/00-15-54.png)
![Screenshot at 21:27: The speaker emphasizes that effective AI adoption should focus on empowering humans rather than simply automating existing tasks to improve employee satisfaction.](https://ss.rapidrecap.app/screens/LToW1rQ6rVA/00-21-27.png)