AI In Your Business

[fisicx]

In engineering there is a difference between inspection and maintenance. But most people use the terms interchangeably. Whilst it will upset those with OCD it doesn’t really matter what you call it. Same with AI and ML. Nobody cares. Except when their subscriptions go up because some AI tools has been foisted on you.

 

[Peter Briggs]

The return question is does your average business owner (user) need or want to know the difference

It's undoubtedly important within the industry, but users need to know benefits and outcomes, not the nuances and technical terminology.

Hi Mark,

Thank you for the response.

I think it is fair that businesses need to know the benefits and outcomes of machine learning to gauge its value, and generally corresponds with the research literature around technology adoption behaviour in business enterprise.

Though, to answer your question, I believe that it is important to make the distinction, mostly because machine learning is a unique subset of artificial intelligence. Technologies such as expert systems, generative AI, and computer vision are based on a machines ability to rationalise by emulating human cognitive functions, whereas machine learning, as the name entails, emulates learning capabilities. The distinguishing factors are that one performs a task (AI) while the other learns to perform or improve at that task (ML).

In a business environment, this is important because it enables many opportunities for improved efficiency and performance, sustainability, innovating business models, quality control, business intelligence, cybersecurity, etc. It is essentially a way to gain a competitive advantage by improving AI-based technologies. However, machine learning requires a resource, which is data. We do, afterall, live in the digital age, and data is the new currency.

 

[Mark T Jones]

Hi Mark,

Thank you for the response.

I think it is fair that businesses need to know the benefits and outcomes of machine learning to gauge its value, and generally corresponds with the research literature around technology adoption behaviour in business enterprise.

Though, to answer your question, I believe that it is important to make the distinction, mostly because machine learning is a unique subset of artificial intelligence. Technologies such as expert systems, generative AI, and computer vision are based on a machines ability to rationalise by emulating human cognitive functions, whereas machine learning, as the name entails, emulates learning capabilities. The distinguishing factors are that one performs a task (AI) while the other learns to perform or improve at that task (ML).

In a business environment, this is important because it enables many opportunities for improved efficiency and performance, sustainability, innovating business models, quality control, business intelligence, cybersecurity, etc. It is essentially a way to gain a competitive advantage by improving AI-based technologies. However, machine learning requires a resource, which is data. We do, afterall, live in the digital age, and data is the new currency.

I'm not undermining your skills or knowledge - I'm pointing out that they are of value to you in making things work / happen, but from a simple user perspective they are actually a distracting from what really matters

How will it make my business better / qicker / more profitable?

It's a failing of many technical businesses (particularly cutting edge tech) that they bury the simple user outcome in mountains of irrelevant technical jargon.

 

[Peter Briggs]

I'm not undermining your skills or knowledge - I'm pointing out that they are of value to you in making things work / happen, but from a simple user perspective they are actually a distracting from what really matters

How will it make my business better / qicker / more profitable?

It's a failing of many technical businesses (particularly cutting edge tech) that they bury the simple user outcome in mountains of irrelevant technical jargon.

Ah, I see. It is tricky to provide examples because machine learning can be applied to many areas in many industries. A simple example, and one that employees fear and dread, for good reason, is the automation of business/manufacturing processes. This alone can streamline your business, reduce costs, and increase production capabilities. Though, I'm very much an advocate of AI technologies complimenting workers rather than replacing them.

If you are interested I can post a bunch of uses for machine learning along with some statistical results that demonstrates its effectiveness?

From what I have learnt during my studies, the power of ML is really in its ability to make predictions (such as predictive maintenance for manufacturers). This is done by learning the operating characteristics of production machines, such as sounds, vibrations, heat etc, through data capture. If you are able to predict machine failure then you save yourself a headache in supply chains, time management, production quality, upsets in materials management/procurement, not meeting demand quotas etc.

 

[Newchodge]

Ah, I see. It is tricky to provide examples because machine learning can be applied to many areas in many industries. A simple example, and one that employees fear and dread, for good reason, is the automation of business/manufacturing processes. This alone can streamline your business, reduce costs, and increase production capabilities. Though, I'm very much an advocate of AI technologies complimenting workers rather than replacing them.

If you are interested I can post a bunch of uses for machine learning along with some statistical results that demonstrates its effectiveness?

From what I have learnt during my studies, the power of ML is really in its ability to make predictions (such as predictive maintenance for manufacturers). This is done by learning the operating characteristics of production machines, such as sounds, vibrations, heat etc, through data capture. If you are able to predict machine failure then you save yourself a headache in supply chains, time management, production quality, upsets in materials management/procurement, not meeting demand quotas etc.

Does it matter to the user what it is called, though?

 

[fisicx]

@Peter Briggs - the end user really doesn’t care. All they want is more profit, simpler processes, lower overheaded etc. The means by which this is achieved is irrelevant. In the same way they don’t care about how their cloud provider manages their data. They just want it to work.

 

[Peter Briggs]

Does it matter to the user what it is called, though?

This is true. It does not matter what it is called. Most people do refer to AI-based technologies simply as AI, and that is understandable. After all, learning is a form of intelligence. However, I would argue that it is important to know the difference. But this ultimately gives me some insight into the mindset of business owners, which was the intention of my original question.

@fisicx Machine learning can provide more profit, simpler processes, and lower overheads. But then comes such questions as how much does it cost? Do I have the employees with the technical expertise to implement and use it? Do I have enough quality data to train AI models? Does it abide by laws and regulations?

 

[ctrlbrk]

I am a PhD student studying machine learning adoption intention in SMEs.

I noticed this thread about AI in business, and I'm curious to know how many of you understand the concepts and benefits of machine learning in business and industry?

Hi Peter,

Like many have said, I don't particularly care about concepts - only benefits.

I frequent a different forum where an AI agent has been deployed and provides answers to, and interacts with users. It is rubbish.

Why? Because so far, all these models carry the now-ubiquitous "this AI model can make mistakes, make sure to check its output" disclaimer.

What good is it to me something I can't rely on? If I have to double check the input every time I ask the question I might as well rely on traditional search engines and other sources.

 

[fisicx]

Machine learning can provide more profit, simpler processes, and lower overheads. But then comes such questions as how much does it cost? Do I have the employees with the technical expertise to implement and use it? Do I have enough quality data to train AI models? Does it abide by laws and regulations?

None of that matters. If I want business improvements I will contract someone to do it for me. I don't need the technical expertise - that's why I pay you.

It's just like IT services. I don't employ a bunch of techs, I contract out this work to experts.

Same with my cleaners. I don't employ cleaners or security guards window cleaners. I outsource.

Remember: I don't want AI or ML or any other cuttingedge tech. All I want is the output from this tech. I tell you my problem, you do some business analysis and provide me with a solution. Which might just be to sack the forklift driver and get someone competent.

 

[Peter Briggs]

None of that matters. If I want business improvements I will contract someone to do it for me. I don't need the technical expertise - that's why I pay you.

It's just like IT services. I don't employ a bunch of techs, I contract out this work to experts.

Same with my cleaners. I don't employ cleaners or security guards window cleaners. I outsource.

Remember: I don't want AI or ML or any other cuttingedge tech. All I want is the output from this tech. I tell you my problem, you do some business analysis and provide me with a solution. Which might just be to sack the forklift driver and get someone competent.

OK. So it's clear that you prefer to outsource a lot of things. My question would be, do you seek to remain competitive, especially in terms of a technological advantage? Or is your management style more reactionary, whereas you would only consider innovation if you felt your business was falling behind in the business environment? What about, lets say, if business partners or active members of your supply chain use certain technologies that would accommodate and streamline your business processes if you also adopted that same technology?

 

[fisicx]

It all depends on the business. If I felt reviewing my processes would help I’d consider getting in a business consultant.

It’s not about the tech. It’s about the processes. Sometimes tech will help, sometimes you just need to change how you do things. Could be as simple as empowering staff to do things without constant supervision or approval.

 

[Mark T Jones]

Benefit- how your product or service meets a specific need as stated by the prospect

Grasp that, and you've grasped selling

Personally,if you've identified a specific need, I dont give a flying ... if you are selling AI, machine learning or off the shelf software - as long as it is the right solution at the right price.

If you go off on tangents lecturing me on the difference between AI and machine learning, my eyes will glaze over and I'll look elsewhere.

Of course, you may be specifically targeting prospects who do care, who are interested in the differencebetween AI and machine learning, in which case your pitch should reflect that.

There is such a thing as too much knowledge in selling.

 

[Peter Briggs]

Ah, this has been interesting. Thank you all for your responses.

My job isn't to judge, criticise, sell anything, or lecture anyone on what they should or shouldn't think or believe. I'm a researcher and social scientist. I simply ask questions or throw out statements and observe the responses. It is what I would do at an AI in Business conference as a way of gauging attitudes toward AI technologies. It is something that research literature is unable to provide, kind of determining which way the wind is blowing, so to speak.

 

[Newchodge]

OK. So it's clear that you prefer to outsource a lot of things. My question would be, do you seek to remain competitive, especially in terms of a technological advantage? Or is your management style more reactionary, whereas you would only consider innovation if you felt your business was falling behind in the business environment? What about, lets say, if business partners or active members of your supply chain use certain technologies that would accommodate and streamline your business processes if you also adopted that same technology?

What a load of rubbish questions. You do not understand how small businesses operate and, unless you do, you cannot help.

 

[Peter Briggs]

What a load of rubbish questions. You do not understand how small businesses operate and, unless you do, you cannot help.

I don't understand the hostility, and it is somewhat rude. I'm not here to help, but if anyone would like advice surrounding AI/ML, I would be more than happy to lend my knowledge. This is a forum and forums are about discussion, which was the intention of my questions.

 

[fisicx]

But you said:

I simply ask questions or throw out statements and observe the responses.

that’s somewhat different than someone wanting to offer help and guidance.

 

[Peter Briggs]

But you said:

that’s somewhat different than someone wanting to offer help and guidance.

It is called being polite, but it is clear that this forum is not for me.

 

[Mark T Jones]

It is called being polite, but it is clear that this forum is not for me.

Apologies (for all) we do get quite a lot of newbies pretending they want to help or do research before launching into sales mode

I'm sure youll appreciate there is a big difference in engagement between people voluntarily attending conferences and random business owners looking for specific outcomes though?

 

[Newchodge]

I don't understand the hostility, and it is somewhat rude.

I apologise, but you did say

I simply ask questions or throw out statements and observe the responses.

 

[Paul Kelly ICHYB]

It doesn't matter what you call it, for most people it is about what does it do/is in it for me?

 

[GLAbusiness]

I have almost 50 year's experience in IT. One thing I have noticed throughout this time is that developers and subject experts view the world through the lens of their specialism. It was true in the early days of spreadsheets and continues into the realm of AI. Even to this day Excel experts will see the solution as Excel based when, in fact, MS Access would be more appropriate.

This can lead to development silos where the question becomes "How can I use my pet technology in the business"

I have always held the view that the most effective solutions come when we view things through the lens of user interface (i.e. business improvement)..

So, as a business I am much less interested in "what can ML do for me" My question would be how can I improve the business using the most appropriate technologies.

Don't get me wrong. We need experts in the different technologies, but they must operate within the purview of "expert generalists"
.

 

[Keynote Speech]

Like it or not, AI is here and is only going to muscle its way more into our lives.

Developers and promoters will tell you that it can do anything and it can do a lot of the mundane stuff, but I still see much of it as extended automation.

If you could have AI do something for you to help your business, what would it be? Remember, that we are not talking about anything physical (YET!).

Something I’ve taken from Cassie Kozyrkov’s thinking is that the real value of AI isn’t magic decision-making, it’s better decision support. She talks a lot about separating prediction from judgment. Machines can help us process vast amounts of information and generate probabilities, but humans still need to define the objective and make the final call.

So if I could use AI for one thing in business, it would be improving clarity around decisions. Summarising data, highlighting likely outcomes, flagging risk patterns. Used properly, it should strengthen human judgement, not replace it.

 

[Mark T Jones]

Something I’ve taken from Cassie Kozyrkov’s thinking is that the real value of AI isn’t magic decision-making, it’s better decision support. She talks a lot about separating prediction from judgment. Machines can help us process vast amounts of information and generate probabilities, but humans still need to define the objective and make the final call.

So if I could use AI for one thing in business, it would be improving clarity around decisions. Summarising data, highlighting likely outcomes, flagging risk patterns. Used properly, it should strengthen human judgement, not replace it.

Personally I'd use it to create lingerie but ultimately meaningless answers

 

[JEREMY HAWKE]

Personally I'd use it to create lingerie but ultimately meaningless answers

It really is quiet at fit out money HQ this afternoon

 

[fisicx]

So if I could use AI for one thing in business, it would be improving clarity around decisions. Summarising data, highlighting likely outcomes, flagging risk patterns. Used properly, it should strengthen human judgement, not replace it.

Except recent research suggests AI is dumbing down summaries. It’s missing nuances and detail and outliers because these aren’t in the LLM. Which can mean poor decisions.

 

[Keynote Speech]

Except recent research suggests AI is dumbing down summaries. It’s missing nuances and detail and outliers because these aren’t in the LLM. Which can mean poor decisions.

That is a fair point!

AI would need to become better at summarising data to ensure accuracy rather than just speed. The danger is when people treat outputs as conclusions rather than inputs. Used properly, it should support scrutiny and better questions, not shortcut thinking.

 

[GLAbusiness]

I recently asked chatGPT to answer some questions about the way it was answering me. It agreed that

• There was a consistent bias towards youthful and flattering defaults when key constraints were missing
• Assumptions were made about the relative importance of the available data, rather than validating which signals actually mattered
• Accuracy improved only after those assumptions were explicitly challenged

 

[Mark T Jones]

It really is quiet at fit out money HQ this afternoon

Just spotted the embarrassing typo - and yes, very quiet!

 

[Newchodge]

Just spotted the embarrassing typo - and yes, very quiet!

i give up, what did you mean to say if 'answers' was a typo?

 

[ctrlbrk]

i give up, what did you mean to say if 'answers' was a typo?

I think 'lingerie' was a typo - probably autocorrect.

 

[Newchodge]

I think 'lingerie' was a typo - probably autocorrect.

i agree. I just thought my interpretation was funny.

 

[tony84]

I have just built a CRM on base44. It will cost me about 20% of what my current CRM does and it does it better. In essence its been designed by us to fit in with OUR process rather than designed for the masses. It has taken nearly 10 pages of instructions. Some bits it got wrong and needed quite a lot of detailed instruction and even then we had to get a human involved to find the problem and fix it. But regardless it is still very impressive.

It reminds me of an English lesson I did about instructions. We had to write the instructions to make a cup of tea. Get a tea bag, put it in the cup... Go to the tea bags, open the container, pick up the tea bag. Walk over to the cup, pick up the cup put the tea bag in the cup... You get the idea.

I am still not completely sold by AI but I can see it will likely get there. You ask a question for example and the answers are not always correct, it will also alter the answer based on how you ask the question.

 

[EnterprisePro]

Anyone making use of AI in their accounting processes

Yes, I do, and it has become a fundamental part of our business process.

The system I’ve developed interfaces with our CRM, Xero and our own internal systems to cross-check data reliability, preserve a single source of truth, support batch handling, and maintain auditability across the workflow.

It also helps with intelligence gathering and process governance, so decisions, handling rules and exceptions can be tracked properly rather than living in someone’s head or getting lost across emails and spreadsheets.

In practice, the real value comes from using it in a controlled way, so repetitive and document-heavy work moves faster without losing consistency or review discipline.