Pollinger AI

Ever stood in front of a washing machine thinking, “What does that flashing symbol actually mean?” 😄

Here’s a practical use for AI: show it what’s happening while you talk through the problem.

In ChatGPT’s phone app, select Advanced under Settings → Voice, start a voice conversation and tap the camera button. Video is available to eligible subscribers, with usage limits.

Point your camera at the controls and try:

“Act as a patient appliance support assistant. My washing machine won’t start. Ask me for the make and model, then help me work through basic checks using the manufacturer’s instructions. Explain things in plain English, one step at a time, and wait for my reply. Start with simple possibilities, such as the child lock. Don’t guess what an unfamiliar symbol means or suggest dismantling anything.”

The useful bit is being able to ask, “Do you mean this button?” while showing it exactly what you’re looking at.

It could help you understand a setting, spot something you’ve overlooked or explain the problem more clearly to a repair service.

AI can make mistakes, so check its advice against the manual and leave electrical repairs to a qualified professional.

Which appliance in your home requires the most negotiation?

4 million steps! 👣

I’ve just passed the 4 million steps mark for 2026.

That’s come from managing at least 12,000 steps every single day, with an average of more than 15,000 a day across the year.

Some of those steps have been training runs, some have been races, plenty have just been everyday walking — but I’m quite pleased with the consistency.

And there’s a purpose behind plenty of the miles at the moment.

On 4 October, I’ll be taking on the Cardiff Half Marathon, raising money for Age Connects Cardiff & the Vale, a charity I’m proud to volunteer with.

So if my 4 million steps are worth a few quid, I’d really appreciate your support:

👉 https://www.justgiving.com/page/jonathan-pollinger-2026

Now onwards towards 5 million…

Marketing people… how much of your week disappears into getting things out the door?

The newsletter. The social posts. The website updates. And that “quick video” someone has just requested.

AI can help with that workload. But it’s worth looking beyond asking it to write another post.

You could use it to:

• Spot recurring questions in customer feedback.
• Turn a webinar into several useful pieces of content.
• Build a campaign brief around your actual budget and capacity.
• Make sense of results and identify what to test next.

My latest blog walks through practical ways to do this, including my Prompting Pillars and Draft, Improve and Polish approach.

Start with one task that regularly eats into your week. Include checking and editing time when deciding whether AI has helped.

Read the blog:
https://www.pollinger.ai/blog/practical-ai-for-busy-marketing-teams

Want the full Marketing AI Cheat Sheet, or some practical training for your team? Get in touch.

AI for Marketing Teams: Practical Uses — Pollinger AI Discover practical ways marketing teams can use AI for research, campaigns, content and reporting, while protecting brand voice and checking results.

Two weeks today I’ll be running 13.1 miles as fast as I can at the Cardiff Half Marathon for Age Connects Cardiff & the Vale. 🏃

Hopefully still smiling at the finish. Almost certainly making some fairly alarming breathing noises. 😅

As a volunteer befriender with Age Connects, I’ve seen how much a regular chat and a friendly face can mean to an older person who might otherwise feel lonely or isolated.

Everyone deserves to feel valued, respected and connected. That’s why I’m aiming to raise £500 to support their work.

A big thank you to everyone who’s already sponsored me. If you can spare a few pounds, I’d really appreciate your support.

I’ll take care of the running!

👉 Please sponsor me on JustGiving:

Jonathan's fundraiser for Age Connects Cardiff and the Vale Help Jonathan Pollinger raise money to support Age Connects Cardiff and the Vale

AI is already at work. Training hasn’t caught up.

A new Deloitte survey of 25,000 UK workers found that 63% are using generative AI for work.

Yet around half have received no formal training on how to use it safely and effectively.

Nearly a third of users are even using AI without their employer knowing.

That tells me businesses don’t have an interest problem. People clearly want to use AI.

They have a support problem.

Without practical training, employees are often left to work things out for themselves. That can mean missed opportunities, inconsistent results and sensitive information being put into unsuitable tools.

Effective AI training needs to go beyond a few clever prompts. People need to understand which tasks AI can genuinely help with, how to provide context, how to check the output and when human judgement must take over.

That’s exactly the kind of practical training I provide through Pollinger AI.

If AI use in your workplace is growing faster than the guidance around it, let’s have a chat.

Survey details: https://www.deloitte.com/uk/en/about/press-room/british-workers-spend-one-billion-pounds-of-their-own-money-on-gen-ai-for-work.html

Got a drawer full of cables you’re keeping “just in case”?

Somewhere in there is probably a lead for a gadget you stopped owning in 2014.

Here’s a small AI experiment: photograph one cable, with both ends clearly visible, and upload it to ChatGPT.

Try this:

“Act as a patient tech helper. I’m sorting old cables at home. Help me identify the connectors in this photo and explain their common uses in plain English, for example connecting a printer. Separate what you can see from what you’re guessing. If you’re unsure, ask for a closer photo.”

The useful bit isn’t just getting a name. It’s knowing what to check next.

Treat the answer as a starting point. A photo can’t establish electrical safety or guarantee compatibility, so check the device instructions before plugging anything in.

What’s the oldest mystery item in your cable drawer? 😄

Halfway there! £250 raised towards my £500 target.

Sadly, that doesn’t mean I can stop halfway round the Cardiff Half. 😆

A big thank you to everyone who’s sponsored me so far. With less than three weeks until race day on 4 October, I’d love to get the remaining £250 in the bag.

I’m running for Age Connects Cardiff & the Vale, where I volunteer as a befriender. I’ve seen first-hand how much a regular chat and a bit of company can mean to someone who might otherwise feel lonely or isolated.

If you can spare a few pounds, I’d really appreciate your support. Sharing this post would help too.

I’ll take care of the running… can you help me over the fundraising finish line? 💪

Donate through my JustGiving page:
https://www.justgiving.com/page/jonathan-pollinger-2026

Your first AI answer is a draft. Treat it like one.

In 2026, we’re used to instant gratification.

Dinner ordered through Uber Eats. A series ready to play on Netflix. Something new every time we open social media.

It’s easy to bring that expectation to AI: type a prompt, get the perfect result, move on.

But AI needs a slightly different mindset. An instant answer isn’t necessarily a finished piece of work.

That’s why I use a simple prompting technique in my training:

DIP: Draft, Improve, Polish.

📝 DRAFT
Get a useful starting point. Explain what you want, who it’s for and give it relevant context.

Don’t expect the right result first time. Give yourself something to work with.

🔧 IMPROVE
Read the response. What’s missing, vague or off target? Give specific feedback.

“The opening is too general. Lead with the customer’s problem and use the example I’ve provided.”

You can also ask AI to suggest improvements, then choose which ones are worth making.

✨ POLISH
Once the substance is right, refine the wording, flow, tone and format.

“Keep the meaning, cut the repetition and make it sound more like me.”

Check the facts along the way. A beautifully written mistake is still a mistake.

It’s been lovely to see other trainers incorporating DIP into their own training. It’s a simple structure that makes the process easier to explain and remember.

Feel free to use it, whether you’re an AI user or a trainer.

Some tasks will need very little revision. Others will take a few rounds. The useful habit is knowing what needs improving and saying so.

Next time an AI response feels “nearly there”, give it a DIP before giving up.

One month to go!

One month from today I take on the Cardiff Half Marathon on Sunday 4 October, raising money for a cause that means a lot to me, Age Connects Cardiff & the Vale.

The challenge is set:

🏃 13.1 miles
⏱️ Sub 1:33 targe
🫴 £500 to raise

I’m proud to volunteer with Age Connects and have seen
first-hand the difference their work makes to older people across our community, particularly those experiencing loneliness and isolation.

So while I’ll be chasing a PB on race day, the more important
target is that £500.

One month of training, whatever the weather! and fundraising to go.

If you’d like to support me, even a few pounds will help make a difference.

Please sponsor me here:
https://www.justgiving.com/page/jonathan-pollinger-2026

And if you can’t donate, a share of this post would be hugely appreciated.

Thank you!

JP

Pic from Bute Park early this morning.