In this podcast from The SEO Works 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.

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!

In this episode from The SEO Works, 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.

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

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.

This year, consumer journeys are messier than ever. People don't always start with Google. They’re picking things up from TikTok, Reddit, or even AI tools, often before they know what they’re really looking for.

Google still leads, but generative AI is evolving how people search. More users are getting answers without clicking anywhere. That means brands need to be visible not just to humans, but to machines that are scanning and summarising.

One thing hasn’t changed: content still matters. It needs to be real, useful and well made. But the focus is shifting. Clear facts aren’t always enough... "vibes" are influencing consumers and people are paying more attention to tone, feeling and style.

We discuss the trends in consumer behaviour shaping 2026.

And while the tech keeps moving, the basics still count. A solid site, strong content, and a clear brand are still what get you seen, even as AI becomes more adopted.

We sum up suggesting the best approach in 2026 is a mulfi-faceted one. Building content that can work across different channels, making it easy for AI to process. But never losing your own uniqueness or forgetting who it’s really for.

This podcast breaks it down for you. No jargon, no waffle... just the key takeaways.

We talk about how ChatGPT has gone mainstream. While plenty of people still use it for work (writing, coding and so on) the majority of messages now are for personal use.

We also explore the kinds of things people are using it for. Writing is still a big one, but there’s also things like multimedia work, asking questions, getting practical advice, solving technical problems, and even using it as a way to think out loud. The percentages will surprise you!

The study also showed how usage varies by age, job and education. There’s a steady shift towards equal use between men and women. And more people are turning to ChatGPT for help and learning, not just to look things up.

Find out how people ACTUALLY use ChatGPT in our this latest podcast from The SEO Works.