Mercia AI - Simplifying AI for You

28/10/2026

AI agents can do a lot — but should they? Discover practical guidance on granting permissions, setting approval gates, and limiting costly mistakes in our latest blog: AI Agents Can Act for You — But What Should You Let Them Do? Read now: https://wix.to/OwBEPng

How much control should an AI agent have? A practical guide to access limits, human approvals and keeping mistakes manageable.

21/10/2026

A MotoGP bike is packed with sensors, but how much of its technology is actually AI? 🏍️

Our new blog explores Ducati’s machine-learning work on tyre behaviour, Aprilia’s AI development programme and why rider feedback still matters when engineers study the data. It also explains why traction control and other sophisticated electronics aren’t automatically artificial intelligence.

There’s a useful lesson beyond racing: better information works best alongside human experience.

Read How AI and Data Are Changing MotoGP:
https://www.merciaai.com/post/how-ai-and-data-are-changing-motogp?utm_source=facebook&utm_medium=organic_social&utm_campaign=blog_how-ai-and-data-are-changing-motogp_2026-10-17&utm_content=facebook_post

23/09/2026

How many AI subscriptions do you actually use?

It is easy to add another monthly plan because a new feature looks useful. The harder question is whether it performs a job your existing tools cannot already handle.

This practical subscription audit looks at overlapping features, free tiers, integrations and the difference between accessing an AI model and needing the complete product built around it.

Before paying for the next tool, ask: what would it let me do that I cannot already do well enough?

Read the full article:
https://www.merciaai.com/post/how-many-ai-subscriptions-do-you-need?utm_source=facebook&utm_medium=organic_social&utm_campaign=blog_how-many-ai-subscriptions-do-you-need_2026-09-19&utm_content=facebook_post

16/09/2026

An AI agent might be able to organise your inbox, update a record or prepare a reply. But how much should you let it do without checking with you?

For a small business, a useful starting point is the least access needed for the job:

• Read and suggest.
• Prepare the work, then ask for approval.
• Act within clear limits once the task has been tested.

Before connecting one, check what it can see, change and send — and whether you can track and undo its actions.

Our new blog walks through these choices in plain English.

Read it here:
https://www.merciaai.com/post/ai-agent-permissions-small-businesses?utm_source=facebook&utm_medium=organic_social&utm_campaign=blog_ai-agent-permissions-small-businesses_2026-09-12&utm_content=facebook_post

09/09/2026

Do you need a new computer to get value from AI, or will your existing equipment and a few cloud tools do the trick? 🤔💻

As someone who’s spent decades building software, I’d start by asking how often you’ll use AI, whether the work is private, and whether your current machine can handle small models. For occasional tasks, cloud tools are usually cheaper. For repeated, private workflows, local or hybrid setups can make sense - but only if the use case justifies the cost.

Read the practical guide and decide with real-world constraints in mind: https://wix.to/LjLYSW8

Which sounds like you - occasional user, privacy-focused, or somewhere in between? Tell me below.

08/09/2026

I’ve put together a plain-English post on how AI is changing cricket - from pitch-side predictions to fan experiences - written from practical, developer-led experience. It’s a quick, 3-minute read that cuts through hype and points to realistic uses for clubs and community groups. 🏏🤖

With England v Pakistan coming up (Third Test, Wed 9-Sun 13 Sep), now’s a good moment to spot tools that add value without breaking the bank. Read it here: https://wix.to/tqUISh8

What’s one part of cricket you’d like to see sensible data or AI applied to at club level? Let me know - I’ll reply.

05/09/2026

New WSL super league starts this weekend - a good moment to look at how AI is changing football at every level. In our latest blog I explain, plainly and practically, where AI is useful (scouting, wearables, set-pieces, broadcasting) and where teams should be cautious. ⚽🤖

Read the post for no-nonsense, founder-led advice and examples you can apply to clubs, community projects or small businesses. https://wix.to/x6z8AfG

What part of AI in sport do you want explained next? I’m happy to dive deeper. 👇

22/08/2026

The 2026/27 Premier League season is underway—and artificial intelligence is becoming part of how the competition is analysed, officiated and experienced.

Clubs now receive tracking data with less than a second of delay. Semi-automated offside technology is being refined, AI assistants are helping supporters explore decades of football history, and the new Fantasy Premier League Companion can provide personalised decision support.

But the most important point is what AI isn’t doing: replacing managers, analysts or match officials.

Our updated article examines where AI is genuinely being used, where deployment is only beginning and why human judgement still matters.

Read it here:

Updated for the 2026/27 season: This article was originally published before the 2025/26 campaign and has been revised to reflect the launch of the FPL Companion, continued development of the Premier League Companion and improvements to semi-automated offside technology.The 2026/2027 Premier League....

22/08/2026

Four AI models made the same prediction before Formula One’s 2026 reset: George Russell would win the Drivers’ Championship.

Eleven Grands Prix later, Kimi Antonelli leads the standings and Russell sits third. But Russell has just won the Dutch Grand Prix Sprint, giving tomorrow’s race a timely new checkpoint for that original forecast.

That does not prove the models right—or wrong.

It demonstrates why predictions should be tested as new evidence arrives, rather than treated as fixed answers.

Our original analysis examines why four different AI models converged on Russell, which assumptions influenced them and what could cause the consensus to fail.

Read the analysis:

https://www.merciaai.com/post/f1-2026-when-four-ai-models-agree-in-a-reset-year?utm_source=facebook&utm_medium=organic_social&utm_campaign=sports_dutch-grand-prix_2026-08-23&utm_content=facebook_post

When several models agree, does that make you more confident—or more curious about what they might all be missing?

Last year, I asked four AI systems to predict the 2025 Formula One World Drivers’ Champion. All four selected Lando Norris. When he went on to win the title — narrowly — the experiment looked like a clean validation of data-driven forecasting under stable regulations.But 2026 is different.The ...

12/08/2026

When I started the Mercia AI blog, I expected to write about AI developments. After reaching 100 posts, those updates evolved into deeper questions about work, local AI, sport, forecasting, infrastructure and community.

These reflections also helped shape Mercia AI itself. It has been a fascinating way to explore how this technology affects our daily lives and local areas.

Read the full reflection:

After 100 Mercia AI blogs, founder Christopher Howell reflects on what local AI, work, sport, forecasting, infrastructure and community taught him about using AI responsibly.