What an academic study can teach us about showing in AI Search

Does AI writing make content more bland?

Are you using AI to write your emails, essays, or reports? You might be losing your own voice and changing your actual opinions without even realising it. In this podcast from The SEO Works [https://www.seoworks.co.uk/] we look at a research paper from Google DeepMind and several universities that investigates how Large Language Models aren't just fixing our grammar, they are fundamentally distorting our intended meaning. Through a combination of controlled human studies and large-scale data analysis, the study found that LLM assistance leads to "linguistic homogenization". The study revealed that even when tasked with minor edits, AI models consistently introduced semantic distortions and replaced individual voices with formal, analytical, and emotional language.  Is the convenience of an AI writing assistant costing us our creativity and personal convictions? We explore the study and the key points that it suggested.

The AI Job Threat: Hype or Reality?

Is AI actually coming for your job, or are the headlines getting ahead of the data? In this episode from The SEO Works [https://www.seoworks.co.uk/], we dive into a groundbreaking 2026 report by Anthropic to separate fact from fiction. We break down their new metric, "observed exposure," which shifts the focus from what AI could theoretically do to what it is actually automating in the workplace today. We uncover why real world AI adoption is currently a fraction of its theoretical potential and reveal the most heavily exposed professions right now. Most importantly, we answer the biggest question on everyone's mind: is AI causing mass layoffs? Join us for a data-driven look at the future of work!

What an academic study can teach us about showing in AI Search

Large Language Models including ChatGPT and Gemini (and even AI Overvies) pull answers from across the web, write their own summaries, and answer questions directly. So where does that leave content creators? In this episode from The SEO Works [https://www.seoworks.co.uk/], we’re looking at a paper from researchers at Princeton, Georgia Tech, and the Allen Institute called Generative Engine Optimization. It is an academic study that explored how content gets picked up and shown in AI generated answers. Here is the paper: https://arxiv.org/pdf/2311.09735 They tested different ways of writing and structuring content to see what actually makes it more visible AI driven engines. Some of the findings are what you might expect. Others are surprising. But all of it points to a shift that marketers and content teams can’t afford to ignore. SEO is not dead at all. But if you want your content to be picked up by generative engines, you may need to also look at things through a slightly different lens. Let's see what this academic study uncovered and how we might apply this in our own content strategies.

How do people ACTUALLY use Claude?

This episode from The SEO Works [https://www.seoworks.co.uk/] looks at Anthropic’s latest research, which studied two million conversations with Claude to understand how people are actually using the AI and what it could mean for the broader economy. The report, called the Anthropic Economic Index, introduces a new way to measure AI’s impact on the economy. It breaks down AI use by task complexity, user skill level, independence, and how often tasks are completed successfully. The data shows that AI helps speed up work that already requires a high level of skill, however there is the chance for more inaccuracy. But it still finds it hard to manage complex tasks that take a long time to complete. There’s a split in how it’s used around the world. In wealthier countries, Claude is often used for group work and collaboration. In less wealthy regions, it’s more commonly used to support learning. The research also raises concerns that AI could reduce the need for skilled jobs by automating them. But the overall effect depends on how reliable the models are, and how widely they’re used in different places. The report is here: https://assets.anthropic.com/m/218c82b858610fac/original/Economic-Index.pdf

What does the UK actually do online?

This podcast from The SEO Works [https://www.seoworks.co.uk/] breaks down Ofcom’s latest Online Nation report, so you don’t have to read through all the pages yourself. We’ve pulled out the key points: who’s using what, how people feel about the internet, and what’s changing. Meta and Google still lead the way. WhatsApp and YouTube are used by nearly every adult who’s online. Most people in the UK are connected, but how they get online varies. Younger people and those with higher incomes tend to go for iPhones. Older users mostly stick with Android. There’s also growing concern about safety online. More people are starting to question whether the internet is doing more harm than good. Children’s use is looked at closely too. Many say they’re happy online, but they’re also staying up late and sometimes seeing things they shouldn’t. We also dig into the rise of AI search tools and how people are responding to age checks being rolled out on adult sites.