TLF Gems Newsletter July 2025

Your monthly CX and insight newsletter from TLF Research

Drawing is essential to understanding form.

Milton Glaser

There are many reasons to be sceptical about the net value of large language models (and a few to be more optimistic, in fairness).

One thing that has relevance outside of the world of AI is just how much quality matters. LLMs are characterised by low to mediocre quality, but output at impressive speed. They've exacerbated the main problem of the digital age, which has been a surfeit of "content" making it harder and harder to sift through to the stuff that actually has value.

Quality still stands out. If you deliver it consistently you will find a market that values it and is prepared to pay for it - that's what a customer loyalty strategy is all about.

Thanks for reading,

Stephen

Here are 6 things we think are worth your time this month


Gopnik on AI

The AI models making all the waves in the news are a dead end that will never deliver AGI, but there are other approaches that just might. In this fascinating lecture, Alison Gopnik talks about how current AI models work and how future AI models could work with a more thoughtful approach to learning, empowerment, and understanding causality. "What it's doing is putting together intelligences of lots of individual people, rather than being an intelligent agent on its own."

Smallest Viable Market

One of the biggest challenges with strategy is figuring out who your market is. Most businesses still tend to define their market as everyone, but Seth Godin makes a very good argument that it's more effective to target a specific piece of the market and grow than it does to target the whole market and hope to fail down to a big chunk of it. "The thing is, you don’t get to 3% of the market by trying for 40% and failing. You get there by embracing the 1% and doing such a good job that the word spreads."

The Work is Never Just The Work

This is a great article by Dave Stewart (about development, but it applies to anything) about why we consistently underestimate how much time and effort everything will take. The categories of "the work before the work", "the work", "the work between the work" etc. are a really useful model to help think about these things. "The key takeaway is that even a detailed estimate of 'the work' can miss significant 'invisible' work."

Learn CPR in 15 Minutes

This is a brilliant idea from the British Heart Foundation. Follow this 15 minute training with your phone and a cushion and you could learn how to save a life on your lunchbreak. "Many of us will witness a cardiac arrest in our lifetime. Be ready for that day with RevivR, our fast, free and easy-to-use online tool."

Cognitive Debt

I really like the term "Cognitive Debt" which John Willshire has coined, by analogy to Technical Debt, to describe the problems we're building up for ourselves by trying to rush to solutions without thinking about where those solutions come from. And yes, I am talking about AI. "Perhaps this is a useful measure for Cognitive Debt; are we incurring a debt by not doing the thinking in the present that we will need to demonstrate in the future?"

What I'm Reading: Not the End of the World

Still making my mind up about this one, to be honest, but it's certainly interesting. The premise is that, whilst there is no doubt that we face a moment of climate crisis, there is more reason for optimism than you might think. This is refreshing, and Ritchie's perspective is based on solid data, but there are some clear pieces of the puzzle that are missing or glossed over. In any case, there's no doubt that we need to act now! "We have the opportunity to be the first generation that leaves the environment in a better state than we found it."