Anyone here taking benefits of AI in their everyday business?
- Thread starter Toprated
- Start date
I always have a ChatGPT tab open and ask it all sorts of questions.
It's especially good at short code snippets such as Linux bash commands, or short snippets of PHP. The longer the files get the more it struggles to do decent work, but it's still incredibly useful and only getting better.
It does occasionally get stuff wrong, and does so confidently with very plausible sounding reasoning.
I'm in the process of installing OpenClaw, which should reduce the copy/pasting between the chat window and the terminal but I don't think it's the magic AI agent that the hype would suggest.
Paul.
- Original Poster
- #3
@antropy Thanks, I also use ChatGPT but in tech work Gemini helps better. Like prompt generation for other AI tools to generate website design. Within a minute, I get the design idea and share for idea approval. But Gemini has been too much speculative. Its doesn't sticks to facts, mixes with older requests with new requests though clear prompt given. But AI is also feeding on us by using our server resources to train their models. Currently I'm working on a Voice AI with face capture, it will give us answers to someone we are talking.
- Original Poster
- #5
Yes, Call center jobs have been badly affected. Companies are increasingly replacing human agents with AI chatbots and automated systems, leading to layoffs.
Given the current state of the technology, it is better to keep a “human in the loop.” Let the system draft the answer, but require a human to approve and send it.
Each correction becomes training data. Instead of training your team again and again, you train the system once and let it scale.
I think that’s where the real challenge is: building a system that learns fast enough to reduce human effort, without losing control or quality.
That is one of the problems, current AI models don't actually learn after they are trained.
Paul.
You’re right. Once those models are trained, that part is basically done.
I was thinking about retrieval-augmented generation (RAG), see entry on wikipedia (https://en.wikipedia.org/wiki/Retrieval-augmented_generation)
A RAG pulls relevant text from databases, uploaded documents, or web sources.
If you build a workflow that feeds corrected answers back into your company’s RAG system, you should create a learning effect over time.
Or better still, just keep people employed and contributing to society.
The ecological and economic damage caused by the ever increasing demands of AI are obscene. The money spent could alleviate many of the world’s problems instead we just make rich people even richer.
It will certainly solve some but will exacerbate others. There is much debate on the value of committing so much resource to tools that so far has not really improved business productivity. I agree AI/ML has huge benefits for science and medicine but equally these tools have generated huge amounts of AI slop. So much so that peer review has now become almost impossible. And yesterday I read that crims have developed AI tools that make system hacking even easier.
Of course it's a double-edged sword. In certain tasks it's helped me immeasurably and has impressed me so much. In others it's just wasted time and come up with nonsense.
What's more important isn't where it's at now, but where it'll be in 5 years.
Paul.
- Original Poster
- #15
From my experience, the free versions of any AI like Gemini is very dangerous. Its speculates and gives answers, even goes beyond to make us happy that we found the solution. And best part is, Pro model updated me on it.
Yes, being "confidently wrong" is one of the huge remaining problems for AI engineers to solve.
Paul.
Jack Dorsey's Block cuts thousands of roles as it embraces AI
The Twitter co-founder says he believes the majority of firms will make similar changes "within the next year."
www.bbc.co.uk
Jack Dorsey's Block cuts thousands of roles as it embraces AI
The Twitter co-founder says he believes the majority of firms will make similar changes "within the next year."
There are pretty consistent outcomes to industrial revolutions (depending who you listen to, this is either number 4 or 5)
- short term spike in job losses
- long term reduction in mid-skill jobs, increase in both low and high skill jobs.
Historically there were also big geographical shifts, which may or may not happen here
If you combine this with the whole WFH thing, I fear there are a lot of mid-skill younger folk actively talking themselves out of jobs.
AI is being touted to take jobs really only by the ones that want to expand their revenue and reduce costs without having people. MS and a couple of businesses already tried it last year and rehired engineers back as it screwed up.
The way that I portray AI is that its like any other tool to enhance a business especially small businesses. If setup right embedded within their processes AI and Automation can deal with the mundane heavy lifting leaving their staff to do the real people work. Things like not needing to manually update CRM systems with the details of their call, drafting the proposals or documentation and emails for the clients so they can move on to the next bit of client work. It should be an extra tool kit that does not add anymore friction to their job.
But to be able to achieve that it requires planning nd understanding of your business processes to then identify where the value is in adding AI and Automation not just piping in claude to your business and expecting it to fix everything. The charlatans have really damaged the space soooo much...
It’s awesome to see someone focusing on AI as a "co-pilot" rather than a threat, especially since it’s such a game-changer for killing off some repetitive tasks. I’ve seen it do wonders for everything from rapid prototyping to deep-diving into customer data that used to take weeks to analyze. The real trick is moving past just using the tools and actually learning how to build them into your business strategy.
I've been using OpenClaw a lot recently with OpenAI's Codex model and I have to say it's been pretty rubbish - always forgetting things you told it, always doing the wrong thing, being stupid and dopey. For the most basic of tasks it's faster than writing the code I guess, but it really does have a long, long way to go.
Paul.
Paul.
So best I have found with Openclaw is to really nail down each workflow you want as a skill so there is not wiggle room for the steps it takes. Also use a knowledge base skill like QMD so you can force it to remember and then also try out lossless claw for memory which basically stores all the chat messages in a sqlite file and does summarisation, compaction etc so you can keep memory across sessions
I used Claude yesterday to build an incident reporting plugin for a volunteer organisation. The plugin works but because Claude is using other people’s code you end up with a whole load of css that needs reworking. You could spend hours crafting the prompt but it’s far quicker to edit the files yourself.
And this is the best way to use generative AI: as a framework which you then adjust for your business.
And this is the best way to use generative AI: as a framework which you then adjust for your business.
Sure its quicker for you but for someone else its probably not. Its easier to get and understand how to get the prompt right and continue to prompt it if its not doing exactly what you want
Not so sure about that. Because I know how to code and how financial calculations work I was able to write the prompt. The client tried and failed because he didn’t really understand how thinks should work.
It’s not about the prompt itself. It’s understanding the requirement. That’s a skill many don’t have.
But many small business owners (possibly most) don't have that understanding and a lot of time would be needed to learn and practice it, if that were even possible. AI proponents tend to be the type of people who think in terms of breaking down processes into their component parts - that is probably why they re into AI. Most people don't think that way.
Small Businesses use experts for marketing, legal advice and outsource their IT infrastructure etc. For understanding AI should that not also be the case? Use someone who does think that way to get the most out of AI for the business and even then just to know processes and automate it without AI too? Sure dont go and get one of the big4 consulting lot who will embed their staff in the business but a proper consultation/process audit to figure out where to start and get on the track with AI literacy just like how Data Literacy is so important in this day and age
That is needed for any process change, certainly not just for AI. Indeed as @fisicx is fond of saying AI is not necessarily the answer. However such an analysis is beyond the financial means of many SME's, who also don't "use experts for marketing, legal advice and outsource their IT infrastructure etc."
Then what is the answer? Either they have to start learning and understanding themselves or they get someone who can help them. A free resource is valuable but it requires time to understand it and there a some courses that again require time to understand. And what sort of price is cost prohibitive if its able to unlock more time and revenue down the line? I saw with alot of small firms when it came to starting to use Data they fell into bad habits of just using a graduate who lacked the knowledge and experience to give them the real value and it could fall the same way of AI just far more dangerous in respect of data protection and leaking customer and company data with those who lack the knowledge using it in the wrong way
Who is shaming them though?
And what happens when their processes price them out of the market? Or they cant keep up with the market around them? Or they cant keep up with regulation changes?
The same argument was made around Data and "digital transformations". Many of those businesses who didnt evolve with those times died off. This isnt saying they should use AI for what they deliver its picking what Automations or AI that can actually help them to work like they are now but quicker and keep up with the ever evolving market. And if you dont know where to start then find someone who does.
This is starting to look like FOMO - and in my view, selling by FOMO usually means they have run out of actual benefits.
My reality here - as a business owner, you need to know where you are in the market and what is going on around you.
If you are in a segment where AI will leave you stranded, you need to act. Otherwise, you can watch & wait.
Personally, with only 5 or so years left, my stance will be to proudly promote not being driven by AI; being a real, experienced person interacting with other real people. Yes, it will exclude me from rapid deals (thank goodness) - it will attract people I want to deal with.
This is the issue that many older generations of business owners face; the landscape of business has changed so mauch in the last 20 years and is continuing to change at a pace that many (including myself) will struggle to keep up with.
The 20 somethings coming into the economy now and doing business typically want to do business in a way that some business owners who have been around for ~20 years don't want.... Ok @Mark T Jones has just posted a reply as I was typing this that makes my point exactly.
There are people like some of us who may not want to change too much, and we're a dying breed. Soon enough, we'll be pushing up daisies, and our business will either be gone or passed on to a younger generation. That younger generation will make decisions they see fit, interact and use tools as they see fit, and it really doesn't matter what the likes of us @fisicx @Mark T Jones or any other 'been around the blocks' think
. The next generation of business owners will want instant results, quick tools, snap decisions, because that will be the norm, and if they cannot operate at that pace, they'll be left behind.
Many small businesses aren’t data driven. They market their services, gather leads and deliver services. They are unlikely to be left behind because they don’t use AI. People will still need boilers installed, roofs fixing, grass cutting, bathrooms refurbished and the million and one other services they use everyday.
Actually, forget AI, I'll use Teams/Zoom as an example that happened yesterday. I was in a discussion with a handful of younger generation talking about meetings, and I explained I like to meet in person and touch the flesh. I was alienated because meeting in person was such a waste of time, when you could fire up Google Meet and be done with it - and building relationships was an inefficient use of time.
To your point @fisicx you just posted, boilers will become more technical just as cars with self diagnosis, grass cutting will be done by robots, bathrooms will be laid out using AI first (an AI architect tool was released to the market a few weeks ago), technology will evolve into every part of life. The types of jobs that existed 20 years ago will fade out, and new jobs will come into the market, and after we are dead and buried, our way of thinking will be written into history as how the past generations used to think.
"What use could this company make of an electrical toy?"— William Orton, President of Western Union, on being offered the patent for the telephone for $100,000 in 1876.
To your point @fisicx you just posted, boilers will become more technical just as cars with self diagnosis, grass cutting will be done by robots, bathrooms will be laid out using AI first (an AI architect tool was released to the market a few weeks ago), technology will evolve into every part of life. The types of jobs that existed 20 years ago will fade out, and new jobs will come into the market, and after we are dead and buried, our way of thinking will be written into history as how the past generations used to think.
"What use could this company make of an electrical toy?"— William Orton, President of Western Union, on being offered the patent for the telephone for $100,000 in 1876.
On the other hand - The second mouse gets the cheese.
There a currently hundreds of small/micro businesses being FOMOd or excited into shoehorning AI into everything - some of which whom will adapt and win, others will be destroyed by their enthusiasm.
I can instantly identify 2 businesses who I have already stopped using, because they replaced real, human interaction with AI (both still promote themselves as 'friendly family businesses').
On a bigger scale, witness the dot.com bubble 2001 - 80% of the big players disappeared - most of the big names remaining were supplying to the internet, not through it.
The internet didn't blink, but the early adopters suffered.
I agree that you can use AI for all sorts of things. @Data Swami seems to be focusing on business data. My point is many small businesses don't have lots of data to be analysed. The bloke who cleans our wheelie bins doesn't need AI to use the pressure washer. His data is pretty much limited to the names and addresses of his customers.
Yes a robot can cut my lawn but that's just a device I buy. Even if my business does lawn cutting with robots as a business owner I don't need AI to drive round and drop it off at the client's house.
Yes a robot can cut my lawn but that's just a device I buy. Even if my business does lawn cutting with robots as a business owner I don't need AI to drive round and drop it off at the client's house.
In the various gold rushes the people getting rich were the suppliers of shovels, pick axes, whiskey and women.
Practically every car these days is powered by an ECU so the tech does keep moving and mechanics continue to have to adapt. Other businesses will have that too as time goes on especially with the rate at which electronics manufacturers are adding things. But every business has an abundance of data sure they might not collect it right now but there is an abundance of it that can be used. There are far too many assumptions in the small business market that data is too complex and expensive to get a handle of and in turn lack of knowledge and understanding of how AI and Automation can help. I know of so many SMEs that get charged ridic amounts by their IT provider for absolute dross data because they dont know what good looks like and Data literacy is so poor. And the same applies to the move into AI too. When you consider all the tools we use on a day to day basis most of them store vast quantities of data and most of them utilise AI in them so whether for bad or for good.
Its not meant to be fomo is just understanding that on the horizon there are things like that coming that impact many markets and its seeing where you can take advantage of those tools without centring the business around AI.
Latest Articles
-
The Most Dangerous Moment in Business Is When Everything Is Going WellThey Saw It Coming. They Did Nothing. The Lessons Every Business Owner...
- Ozzy
- Updated:
-
AI and Your Business: Useful Tool, Not a Magic WandThere is no shortage of noise around artificial intelligence right now...
- Ozzy
- Updated:
-
Companies House WebFiling security incident: what happened, what to check, and what to do nextCompanies House has confirmed that its WebFiling service suffered a...
- Ozzy
- Updated:
-
Spring Budget 2026: A Quiet Update for UK BusinessesChancellor Rachel Reeves delivered the UK’s Spring Statement on 3 March...
- Ozzy
- Updated:
-
Practical Ways to Make Your Small Business More Resilient (Especially When the Unexpected Happens)Running a business means juggling plenty of things you can plan for and...- UKBF Editor
- Updated:
