Skip to content
3 min read Newsletters

Updates: GenAI Adoption Shifts, Chinese Models, etc.

S-curve adoption stalls, benchmarks churn, and cheap Chinese/open models surge as capital-intensive moats erode across the AI stack.

Updates: GenAI Adoption Shifts, Chinese Models, etc.
Photo by Roman Kraft / Unsplash
💡
Correction:
There was a typo in the original emailed version of this post where I called Gemini a Microsoft model. It isn't, obviously. It's Google's.

A few quick things worth thinking about on a busy Monday morning:

  1. GenAI adoptions changing S-curve
  2. LLM benchmark trends
  3. Chinese LLMs' growing share

Let's do this:

1. GenAI Adoption's Changing S-Curve

S&P has out a new report on GenAI adoption, and, while interesting, it doesn't entirely say what it thinks it says. (As a related aside, I increasingly find the expression "GenAI" as wearying and dated as I once did "high tech", so keep that in mind.)

Here is the main graph that caught my eye:

While the growth is dramatic, the history of technology adoption shows that these S-curves often change quickly after about 50% of the market has adopted. Future gains, while material, are generally slower, more expensive, and often come with higher churn.