AI as the New Business Assistant
What Could AI Actually Do for a Small Business?
AI is often described as replacing jobs. I am more interested in whether it can remove the jobs I do not particularly want to do.
Whenever artificial intelligence appears in the news, the discussion very quickly seems to become dramatic.
Which jobs will disappear?
Will AI replace teachers?
Will it replace photographers?
Will it write films?
Will companies eventually be run almost entirely by machines?
Those are interesting questions, but they are not necessarily the questions that matter most to somebody running a small business today.
For me, the much more practical question is:
What work can AI take off my desk?
That is rather different.
I do not particularly want an artificial intelligence system to replace the parts of my work that I enjoy. I enjoy teaching. I enjoy experimenting with photography. I enjoy making films, designing products, sailing, working with equipment and talking to customers.
What I would quite happily surrender are some of the repetitive, administrative and time-consuming tasks that surround those activities.
That is where AI becomes interesting.
Philip M Russell Ltd is perhaps a useful case study because it is not a business that does just one thing.
Under one roof I might be:
teaching GCSE or A-level students;
preparing science practicals;
writing educational material;
planning photography;
producing video;
designing personalised clothing or products;
researching equipment;
writing blog articles;
producing social-media material;
organising sailing activities;
answering enquiries;
planning future projects.
Traditionally, many of those jobs would have required either a considerable amount of my own time or several different assistants.
AI is beginning to change that.
Not by becoming the business owner.
But by becoming something surprisingly useful:
a business assistant that is available whenever I need it.
The Small-Business Problem: There Is Always Another Job
Large companies divide work between departments.
There might be separate teams for:
marketing;
accounts;
human resources;
customer service;
research;
IT;
design;
sales;
administration.
A small business rarely has that luxury.
Quite often the same person does everything.
You may spend the morning delivering the service that actually earns the money and the afternoon completing all the tasks required to make tomorrow possible.
There are emails to answer.
Invoices to prepare.
Customers to follow up.
Social-media posts to write.
Websites to update.
Ideas to research.
Photographs to process.
Documents to organise.
Marketing campaigns to plan.
Meetings to arrange.
And somewhere amongst all this, you are supposed to develop the business.
That is one reason AI interests me.
It effectively offers another pair of hands — although perhaps another pair of digital hands is a better description.
AI Does Not Have to Run the Business to Be Useful
There is sometimes an assumption that AI has to perform an entire job before it becomes worthwhile.
I think that misses the point.
Suppose a particular administrative task normally takes 30 minutes.
AI reduces it to 10 minutes.
That may not sound revolutionary.
But suppose the same thing happens with ten different tasks during the week.
Suddenly several hours have been recovered.
For a small business, that matters enormously.
The important measure therefore may not be:
"Can AI do my job?"
It may be:
"Can AI remove 10 minutes from twenty jobs?"
That is a much more interesting proposition.
1. AI as a Writing Assistant
Writing is probably one of the most obvious uses.
A small company produces an extraordinary amount of written material.
Emails.
Quotations.
Website pages.
Blog articles.
Product descriptions.
Social-media posts.
Instructions.
Customer FAQs.
Lesson notes.
Reports.
Promotional material.
The difficulty is not necessarily that business owners cannot write.
The difficulty is finding the time.
I might have an idea for an article while teaching or while working in the laboratory.
For example:
"Why does the wind sometimes appear to die when rain starts?"
That thought can become the seed of an article.
Instead of beginning with a blank page, I can give AI the idea, explain what I have observed and ask it to help organise the article.
I still decide whether the explanation makes sense.
I still add my own experience.
I still change anything that does not sound like me.
But the blank page has disappeared.
That is a significant improvement.
2. One Idea Can Become Several Pieces of Content
This becomes even more valuable when content needs to appear in several places.
Suppose I write a 1,500-word article about a science experiment.
Previously I might then need to create:
a short Facebook post;
a LinkedIn version;
an X post;
a website introduction;
perhaps a script for a short video.
That is five pieces of writing based around the same idea.
AI can help transform the original material into each format.
The key point is that the central idea is still mine.
AI is helping with the packaging.
For a small business attempting to maintain a regular online presence, that can save a considerable amount of time.
3. AI as a Research Assistant
Another powerful use is preliminary research.
Notice that I say preliminary research.
AI should not automatically be treated as the final authority.
It can make mistakes, misunderstand a question or present information that needs checking.
But it is extremely useful for helping identify what needs investigating.
Suppose I am considering buying a piece of photographic equipment.
Instead of beginning with dozens of product pages, I could ask AI:
"What factors should I compare when choosing between these cameras for multi-camera video production?"
It might suggest considering:
recording limits;
autofocus;
overheating;
HDMI output;
lens compatibility;
battery life;
low-light performance;
timecode;
stabilisation;
audio inputs.
That immediately gives me a framework.
I can then investigate those factors properly.
AI has not made the purchasing decision.
It has helped me ask better questions.
That distinction is important.
4. AI as a Planning Assistant
Planning is another area where AI can be extremely useful.
Suppose I want to create a six-week series of science articles.
The traditional approach might begin with me sitting in front of a blank document attempting to think of six subjects.
Instead, I can ask AI for 30 possibilities.
I might reject 20.
Modify five.
Combine three.
And discover two ideas that had not occurred to me.
The result is still my series.
But AI has become a brainstorming partner.
The same approach works for:
lesson planning;
filming schedules;
photography projects;
marketing campaigns;
product ranges;
blog series;
club activities;
equipment testing;
business development.
The advantage is not necessarily that AI has better ideas.
The advantage is that it can generate a large number of starting points extremely quickly.
5. AI in Teaching
Teaching provides particularly interesting examples.
I might be preparing a GCSE mathematics student for an examination.
I can ask AI to create a graded set of questions beginning with straightforward examples and gradually increasing in difficulty.
Or I might say:
"Create five questions testing surds, followed by five involving harder rationalisation, and finish with three Grade 8/9 problems."
That can save preparation time.
But the teacher remains essential.
I still need to know whether the questions are appropriate.
I need to spot errors.
I need to understand the student.
I need to decide which questions to use.
Most importantly, I need to recognise why the student is struggling.
AI can create ten algebra questions.
It cannot automatically know that a particular student becomes anxious when too much material appears on a page or that another understands the mathematics perfectly but misreads examination questions.
That is the human part of teaching.
And it matters.
6. AI as a Teaching Resource Generator
The same principle applies to educational resources.
A teacher could ask AI to produce:
revision questions;
model answers;
lesson starters;
extension tasks;
comprehension exercises;
practical-investigation ideas;
vocabulary lists;
quizzes;
alternative explanations.
One particularly useful technique is asking for the same concept to be explained at several levels.
For example:
"Explain electromagnetic induction for a GCSE student."
Then:
"Explain the same idea for an A-level student."
Then:
"Now explain what additional physics would be studied at university."
That immediately provides a layered teaching resource.
Again, the teacher decides what is useful.
AI accelerates the preparation.
7. AI in Photography
Photography may appear to be an unusual place for AI business assistance, but there are many possibilities.
Before a shoot AI could help produce:
shot lists;
equipment checklists;
lighting plans;
location questions;
customer questionnaires;
pose suggestions;
backup plans.
Imagine photographing an event.
I could provide the timetable and ask AI to create a shot list identifying the moments that must not be missed.
That does not take the photographs.
It makes it less likely that something important gets forgotten.
8. AI in Film and Video Production
Film production involves even more planning.
There may be:
scripts;
storyboards;
camera plans;
sound requirements;
lighting;
interview questions;
running orders;
graphics;
titles;
captions.
AI can assist with almost all of these.
Suppose I am filming an interview.
I might give AI some background information and ask it to generate 25 possible questions.
I probably would not use all 25.
Perhaps only eight are genuinely useful.
But one of those eight might be a question I had not considered.
That alone can make the exercise worthwhile.
9. AI Can Help Before the Cameras Are Switched On
The most expensive part of filming is often not editing.
It is discovering during editing that you failed to record something.
Planning therefore matters.
Imagine a three-camera interview.
AI could help generate a checklist:
Camera 1 — wide shot.
Camera 2 — interviewer.
Camera 3 — guest close-up.
Check white balance.
Check microphones.
Record room tone.
Record cutaways.
Take still photographs.
Record an introduction.
Record an alternative ending.
The list itself is not sophisticated.
But forgetting any one of those things might be inconvenient.
Sometimes AI's value comes from remembering boring things reliably.
10. AI in Personalised Products and Clothing
Another part of my business involves equipment capable of producing personalised items.
This introduces a completely different workflow.
A customer may arrive with an idea that is not yet a finished design.
They might say:
"We need polo shirts for our club."
That produces dozens of questions.
What colour?
Where should the logo go?
How large should it be?
Embroidery or printing?
Should names be included?
What information should go on the back?
What sizes are required?
What happens if the logo is too complicated for embroidery?
AI can help convert an informal idea into a structured production brief.
That means fewer forgotten details.
And fewer forgotten details usually mean fewer mistakes.
11. AI as a Customer Communication Assistant
Customer communication is another obvious application.
Most businesses receive variations of the same questions repeatedly.
"How much does it cost?"
"When are you available?"
"What do I need to bring?"
"How long will it take?"
"Can you do this particular job?"
AI can help draft clear responses.
But this is also an area where businesses need to be careful.
Customers should not feel as though they are speaking to a machine that has no understanding of their situation.
I think the best use is often AI-assisted communication rather than completely automated communication.
AI drafts.
The business owner reviews.
The customer receives a useful personal response.
12. Turning Rough Notes into Professional Communication
This can be extremely useful when time is short.
I might write:
"Can do Tuesday 4pm. Bring calculator. £60. We will look at mechanics first."
That contains all the information.
It is not necessarily how I would want to send it to a new customer.
AI can transform those notes into a polite email in seconds.
The information remains mine.
The presentation improves.
13. AI as a Marketing Assistant
Marketing is one of those activities that small businesses know they should do more regularly.
The problem is that marketing frequently loses the competition for time.
Customers come first.
Teaching comes first.
Filming comes first.
Producing orders comes first.
Marketing happens when somebody remembers.
AI can help maintain a more consistent flow.
For example, one article might produce:
a website blog;
a LinkedIn post;
an X post;
a Facebook post;
a short-video script;
a list of image ideas.
That means a single piece of expertise can be used much more effectively.
14. Marketing Without Sounding Like Marketing
There is another advantage.
The strongest small-business marketing often does not look like an advertisement.
A photographer can explain how focal length changes perspective.
A tutor can explain why students lose marks on longer examination questions.
A filmmaker can discuss microphone placement.
A sailing instructor can explain how to recognise a gust approaching across the water.
Each article demonstrates expertise.
It gives the reader something useful.
AI can help turn everyday professional knowledge into publishable material.
That can be much more effective than repeatedly saying:
"Please buy my services."
15. AI as a Data Assistant
Small businesses also accumulate data.
Enquiries.
Bookings.
Sales.
Website visits.
Customer types.
Products.
Advertising results.
Student marks.
Equipment costs.
Many businesses possess useful information without ever analysing it properly.
AI can help identify patterns.
For example:
Which services generate the most enquiries?
Which marketing posts produce engagement?
Which months are busiest?
Which products are profitable?
Which customer questions occur repeatedly?
Which subjects are students struggling with?
Once again, the AI does not have to make the decision.
It can help reveal the information on which decisions are based.
16. AI and Administration
Administration may be the least glamorous application of AI.
It may also be one of the most valuable.
Consider some of the jobs involved in running even a modest business:
meeting notes;
task lists;
email summaries;
appointment preparation;
document organisation;
customer follow-up;
stock lists;
equipment inventories;
project planning;
reminders.
None of these is particularly exciting.
All of them consume time.
This is precisely the type of work I would like AI to absorb.
17. AI as an Idea Generator
One of my favourite uses of AI is simply asking:
"What have I not thought about?"
This is surprisingly powerful.
Suppose I am planning a photography course.
I might already have:
exposure;
aperture;
shutter speed;
ISO;
lenses;
composition.
Ask AI for additional topics and it may suggest:
metadata;
backup strategy;
colour management;
printing;
weather protection;
location permissions;
workflow automation.
I may reject most of the suggestions.
But one overlooked idea can improve the project.
AI does not need to be right every time to be useful.
18. AI Can Connect Different Parts of a Business
This may eventually prove to be one of the most interesting applications.
Businesses often store information in separate places.
Emails contain customer requests.
Calendars contain appointments.
Spreadsheets contain figures.
Documents contain plans.
Websites contain marketing material.
AI increasingly has the potential to work across those systems.
Imagine asking:
"What customer enquiries have not yet received a response?"
Or:
"What jobs are scheduled this week, and what equipment will I need?"
Or:
"Which customers asked about A-level chemistry during the past three months?"
Or:
"Create a draft social-media schedule from the articles published this month."
That is no longer simply text generation.
It starts to look much more like an actual assistant.
19. But Should AI Be Allowed to Do Everything?
No.
There are obvious limits.
I would not want AI making important customer promises without oversight.
I would not want it automatically publishing technical information I had not checked.
I would not want confidential information treated carelessly.
I would not allow an AI-generated scientific explanation to reach a student without checking it.
I would not trust an AI-generated quotation without confirming the numbers.
And I certainly would not allow an AI system to make important business decisions merely because its answer sounded confident.
That final point is important.
AI can be extraordinarily persuasive even when it is wrong.
Human judgement remains essential.
20. AI Is Very Good at Producing a First Draft
One useful way of thinking about AI is this:
AI is frequently excellent at version 1.0.
Humans still need to produce version 2.0.
The first draft might be:
a lesson;
an email;
a marketing idea;
a script;
a checklist;
a business plan;
a research summary.
My job is then to ask:
Is it correct?
Is it appropriate?
Does it sound like us?
What is missing?
What should be removed?
What would make it better?
That is a much more realistic model than imagining that pressing a button automatically produces the finished result.
21. The Human Becomes the Editor
This may be one of the biggest changes AI brings to small businesses.
Instead of creating everything from nothing, we increasingly become editors.
The AI produces possibilities.
We choose.
The AI produces a structure.
We modify it.
The AI produces a draft.
We improve it.
The AI finds patterns.
We interpret them.
The quality of the final result therefore still depends heavily on the judgement of the person using the technology.
22. Experience May Become More Valuable, Not Less
There is a curious consequence here.
The better you understand your subject, the easier it is to recognise whether the AI has produced something worthwhile.
An experienced photographer knows when a proposed lighting setup is unrealistic.
An experienced teacher knows when a question is badly constructed.
An experienced filmmaker knows when a shot list is missing something.
An experienced sailor knows that a neat theoretical answer may not reflect what actually happens on a particular stretch of river.
AI therefore does not necessarily make expertise irrelevant.
It can make expertise even more useful because expertise allows you to judge its output.
23. What Happens to the Time We Save?
This may ultimately be the most important question.
Suppose AI saves me five hours each week.
What should happen to those five hours?
I could use them to:
teach another student;
develop a new practical experiment;
improve a video;
photograph something properly;
learn new software;
design a product;
go sailing;
research a new service;
or simply finish work earlier.
That is where the real business value lies.
Saving time is not particularly useful unless we deliberately decide what to do with the time that has been recovered.
24. The Goal Is Not an AI Business
I do not particularly want Philip M Russell Ltd to become an "AI company".
I want it to remain a business built around teaching, photography, filmmaking, practical science, technology, personalised products and other activities that interest me.
AI is simply another tool.
A very powerful tool, certainly.
But still a tool.
Much as computers changed how we prepare documents, digital cameras changed photography and nonlinear editing transformed filmmaking, AI may gradually become part of the normal business toolkit.
Eventually we may stop talking about "using AI" altogether.
We will simply use it.
From Assistant to Something More
At the moment, one of the most useful ways to think about AI is as an assistant.
It can help write.
Research.
Plan.
Organise.
Analyse.
Summarise.
Brainstorm.
Prepare.
But the next stage is even more interesting.
What happens when the assistant can not only suggest what needs doing, but can actually carry out carefully authorised tasks?
Could it prepare customer follow-ups?
Could it organise schedules?
Could it examine sales figures each week?
Could it maintain a marketing calendar?
Could it identify unanswered enquiries?
Could it help coordinate an increasingly complicated business?
That is where this series is heading.
Because the interesting question may no longer be:
"Can AI write something for my business?"
It may soon become:
"How much of the machinery surrounding the business can AI quietly operate while I concentrate on doing the work that actually requires me?"
And for a small business owner, that may be a much more important revolution than replacing anybody's job.
Conclusion — Give AI the Jobs You Do Not Want
Much of the public discussion around artificial intelligence begins with fear that AI will replace people.
For a small business, I think there is another way to look at it.
Do not begin by asking which person AI could replace.
Ask which repetitive task it could remove.
Ask which blank page it could fill.
Ask which checklist it could prepare.
Ask which pile of information it could summarise.
Ask which forgotten administrative job it could remind you about.
Then keep the part that matters most:
the judgement;
the expertise;
the relationship with the customer;
the creativity;
the decision making;
and the enjoyment of actually doing the work.
I am not particularly interested in an AI that replaces the interesting parts of my business.
I am very interested in one that takes care of some of the boring parts.
And that may turn out to be exactly what the small-business owner has always needed.
