AI-Wiki

Creating AI Value as Adoption Grows Rapidly

3 min read

Rendtorff-Smith & Chin (BCG) — creating AI value as adoption grows

▶ Watch on YouTube · Boston Consulting Group · 5:27

Smart AI development policies focused on society and the economy help advance innovation. The OECD’s Sara Rendtorff-Smith and BCG’s Vincent Chin discuss how.

TL;DR

A five-minute interview on the Boston Consulting Group channel. Vincent Chin of BCG talks with Sara Rendtorff-Smith, who leads AI and emerging technologies at the OECD. It holds the wiki’s most compact set of official figures on the adoption gap by firm size. Three claims:

  1. Safety and governance as enablers. The tension between innovating and being safe is “often presented as a tension,” but “safety, governance, policy is really an enabler for rolling out the technology.” The OECD recommends no single approach; around 80 countries have AI strategies, and its repository holds about 2,000 government documents.
  2. Money and use have both more than doubled. VC funding for AI was about $260 billion in 2025, 61% of all VC globally, about double 2022. Firms using AI went from 8.7% to 20% between 2023 and 2025.
  3. The gap is below the surface. More than 50% of big firms use AI against about 17% of smaller firms. “It’s not just having access to the technology”: the bottlenecks are skills, data and “the organizational and managerial processes for integrating.” Then “can you absorb it? As you absorb it, what value do you create?”

Key claims

  • Keeping pace. Asked how to regulate something moving faster than policymakers’ comprehension, she describes an iterative approach. Sandboxing is “a nice balance,” and Singapore’s voluntary guidelines show another route. Partners across industry and academia bring the OECD the latest developments.
  • Value chain. Three levels (firm, economy, society) read as a chain: enablers and investment, then diffusion, then value. Of value she says only “there’s still upside.”
  • Membership. 38 member countries, and a partnership of 47 working on AI.

Dynamic-capabilities reading

  • digital-transforming/improving-digital-maturity: skills, data and managerial processes named as what firms lack in order to absorb AI, measured by firm size.
  • contextual/internal-barriers: the bottlenecks that keep smaller firms from using AI once access is no longer the constraint.

How it connects

  • Schwaeke et al. — a systematic review of AI adoption in small firms.
  • Beauchene (BCG) — “Adoption is an individual metric. Value creation…”
  • McKinsey’s CEO agenda — capital expenditure at a larger scale than this VC figure, and the question of when enterprises capture value.

What was actually ingested

The whole video (5:27, length_seconds: 327) from BCG’s human-written captions via the yt-dlp fallback, 28 lines. The captions spell her name Sarah; the description’s Sara is used here. No stills: a frame probe showed an interview. The figures are spoken without a source or year for the firm-size split.

Linked entities and concepts

Dangling (single-source mention, deferred): OECD, Sara Rendtorff-Smith, Vincent Chin.

Scope and reliability

A short promotional interview, but the speaker runs the relevant OECD work, so the figures carry more weight than the format. They are not cited to a named dataset, and “firms” is undefined (OECD series usually cover firms above a size threshold in member countries, so the 20% should not be read as all firms worldwide). She dates the OECD’s AI principles to “around 10 years ago”; they were adopted in 2019.

Relationships

Supports

Graph View

Backlinks