AI & Books: notes from a moral panic
The practice of AI companies using old books for training data provides a case study in how dis- and mis- information spread.
Peter Gasston
Technologist, Innovation Lead, Speaker, Author
Latest posts
How I use AI in creating my newsletter
For a few years now I ve written a weekly newsletter, Tech Landscape, that rounds up new feature and product announcements in digital technology (90% AI these days), social media, and online culture, with some comment and insight where appropriate. It s broadly aimed at people who work in or around digital marketing, or just want to [ ]
Taste is the bar of quality in the absence of correctness
In math, you know 2+2 is 4. But how do you judge a great article? A great pitch deck? A great painting? A great design? Much harder. So much harder that people have boiled it down, for lack of a better term, to 'taste.' But taste is nothing more than a rule of thumb: compressed [ ]
The Future of AI Images
Advancements in AI image generation, particularly through Google's Nano Banana Pro, highlight the need for models to understand causality and physics. Current models struggle with sequences, failing to accurately depict actions like a falling wine glass. Future breakthroughs may involve world models that comprehend physical laws, enhancing image realism.
(untitled)
“We need to create new IP. Culture creates new ideas downstream. Without new IP, it’s like trying to feed yourself by eating your own arm. The past is a foreign country that we should impose tariffs on.”I actually love this thought by Matt Webb; let's radically reform copyright. Shorter terms, stronger enforcement. Encourage new things [ ]
(untitled)
Narrative violation: The approximately 100 occupations most exposed to AI automation are actually outperforming the rest of the [US] labor market in terms of job growth and real wage increases. This suggests that current AI systems are generally enhancing worker productivity and shifting workers’ tasks toward higher-value activities. From Vanguard [PDF] via TechCrunch.
(untitled)
An emerging theme this year is capability overhang — the 'gaps between what the models can do and what most people actually do with them'. OpenAI (on X): Progress towards AGI will depend as much on helping people use AI well, in ways that directly benefit them as on progress in frontier models themselves. 2026 will [ ]
(untitled)
Adam Mosseri, Head of Instagram (on Threads / Instagram): Looking forward to 2026, one new significant shift is that authenticity is becoming infinitely reproducible. Everything that made creators matter—the ability to be real, to connect, to have a voice that couldn’t be faked—is now suddenly accessible to anyone with the right tools. [This] will in turn [ ]
Is JSON prompting effective? (No.)
I noticed a trend of people saying that to get better results from Google Veo you should write prompts using more complex, machine-optimised (JSON) formatting. This didn't seem right to me, so I tested it.
Content, community, and coincidence
These two stories, saved next to each other in my feed reader, are directly related: From Content and Community, by Ben Thompson: So, are existing publishers doomed? Well by-and-large yes, but that s because they have been doomed for a long time. People using Al instead of Google or Google using Al to provide answers above [ ]