The Advantage of AI

LongLensPhotography

Th..th..that's all folks!
Suspended / Banned
Messages
18,695
Name
LongLensPhotography
Edit My Images
No
Ai may sound like and doom and gloom and it is sure wrecking havoc on my traditional roles, not just photography and graphic design.

On the other hand it is a massive opportunity and incredible asset available for free for basic use, or fairly nominal cost.

So, what application have you found for it in your business and photography?

I will start with mine - contract review and drafting, WordPress troubleshooting and css code generation, client comms (saves time) in select cases, strategy planning (better than nothing or BAD advice from someone not in the game or failing). I tend to stay away from editing using AI, but gen fill can be useful for QUICKLY eliminating small imperfections, or trash cans I may have missed or couldn't face to touch with my hands. I should be starting some own youtube work, and music gen sounds pretty tempting!
 
I'm sticking my feet firmly in the mud for now and trying to avoid it wherever possible, but my wife uses it at work for help with wording emails for team updates/course invites etc. She checks and tweaks everything that she sends out though so that it still sounds like her. Some of her colleagues have fallen into the trap of letting AI do all the work though, and she says that it's obvious when people haven't checked things properly, or just generated some text and sent it straight out.
 
This reminds me of "The Last One" (later known as "TLO") which claimed it would allow non-programmers to write their own applications.

Well, "up to a point Lord Copper".

:tumbleweed:
 
In my job, I am an IT consultant who specialises in one particular piece of software. It's a broad piece of software and impossible to know all answers to questions about it's functionality, so I sometimes give AI a go at answering the... and it completely makes up convincing but totally wrong answers!

Ah, but that's not unique to AI. ;)
 
however 90% of image generation by Ai is AI-slop
 
Ah, but that's not unique to AI. ;)
When I attended business conference at nec last year, the speaker in qa session was answering questions on ai. Then came the inevitable one - what about all the hallucinations? He reframed the question back about the employees and co-workers - are they ever wrong, do they completely make up the stuff up? It was obvious the whole audience of very high level people nodded, and the shift was obvious: Ai can be more reliable, and only God is truly perfect.
The right prompts and curation are essential - garbage in _ garbage out. This works with human employees too
 
the whole audience of very high level people nodded...
Define "very high level people" - do you mean designers, programmers and testers or do you mean "managers".

Yes, techies make mistakes as do digital systems. The difference is that programmed systems can't fix their mistakes, because they don't know that they've made a mistake, but techies do know and can fix their mistakes, provided the money men and "managers" give them a chance to do so.
 
Define "very high level people" - do you mean designers, programmers and testers or do you mean "managers".

Yes, techies make mistakes as do digital systems. The difference is that programmed systems can't fix their mistakes, because they don't know that they've made a mistake, but techies do know and can fix their mistakes, provided the money men and "managers" give them a chance to do so.
Not only that, but very often it is assumed that humans will make mistakes, so work is double checked (proofreading, code reviews, etc) while 'AI' appears to often be taken as an 'authority' that does not need checking.
 
Define "very high level people" - do you mean designers, programmers and testers or do you mean "managers".
Business conference would immediately suggest businesses owners, many way above our profits

Not only that, but very often it is assumed that humans will make mistakes, so work is double checked (proofreading, code reviews, etc) while 'AI' appears to often be taken as an 'authority' that does not need checking.
Yes, thats why all trainings and presentations stress the requirement of output verification and if necessary tweak, regardless if ai or human authored it
 
Business conference would immediately suggest businesses owners, many way above our profits
Ah, people highly skilled at getting far more than they deserve, also known as Donnie Trump's funding machine,... ;)
 
I am trying to clarify my insurance policy valuation, and frankly the sort of responses are getting are way below chatgpt 3.0 level (2023)
Meet the real human bureaucracy. Talk a lot, answer nothing
 
Ah, people highly skilled at getting far more than they deserve, also known as Donnie Trump's funding machine,... ;)
Lol, keep saying what you already are to make sure you never get more than you deserve
 
Lol, keep saying what you already are to make sure you never get more than you deserve
I've already got as much as I need.

My problem is with those who have far more than they need and still want more. Gordon Gekko may have been a fictional character but his real life counterparts are even nastier.
 
I've already got as much as I need.

My problem is with those who have far more than they need and still want more. Gordon Gekko may have been a fictional character but his real life counterparts are even nastier.
I would agree 100% with this. Greed is a cardinal sin that afflicts so many.
I retired at 52 when I had enough money and was fed up of the politics and backstabbing!
 
Lots of cameras already use AI built it. I'm told most of the eye tracking autofocus systems use it, some of the built in noise reduction do, and various other things. Love it or hate it, it's here to stay.
I know the purists will hate it, but then people said the wet collodion process was the bees knees.... ;)
 
  • Like
Reactions: nog
I mainly use AI (chatGPT) for helping to draft better/more professional letters and emails. I also use it for research, and have used it for a bit of fun with images.
 
I mainly use AI (chatGPT) for helping to draft better/more professional letters and emails. I also use it for research, and have used it for a bit of fun with images.

My observation is that ChatGPT has devalued these sorts of letters and the commentaries or narratives that are submitted along with other materials.

I find AI for assisting image editing can be useful.

But ....

At work my younger colleagues are still failing to convince me that their coding efficiency and documentation quality is actually better using AI -- or that it saves money in any way at all. My view is that the subscriptions now being paid for AI tools are not actually generating a real return so in reality our net costs have gone up a bit. However - at work that makes me appear increasingly obsolete and curmudgeonly and I'm being pushed back from having any real influence on projects as the next generation take over.

I was watching one of friends venting his frustrations with the accounting software his company uses - which has been 'upgraded' with AI. His view is that it delivers no benefit on the mainstream tasks and has led to some time wasted dealing with unpicking the results of its foibles on some of the transactions. So again - net cost to his company as his time has been consumed - and of course the supplier of the accounting software will probably increase their charges because of the suppoed benefits.
 
My observation is that ChatGPT has devalued these sorts of letters and the commentaries or narratives that are submitted along with other materials.

I find AI for assisting image editing can be useful.

But ....

At work my younger colleagues are still failing to convince me that their coding efficiency and documentation quality is actually better using AI -- or that it saves money in any way at all. My view is that the subscriptions now being paid for AI tools are not actually generating a real return so in reality our net costs have gone up a bit. However - at work that makes me appear increasingly obsolete and curmudgeonly and I'm being pushed back from having any real influence on projects as the next generation take over.

I was watching one of friends venting his frustrations with the accounting software his company uses - which has been 'upgraded' with AI. His view is that it delivers no benefit on the mainstream tasks and has led to some time wasted dealing with unpicking the results of its foibles on some of the transactions. So again - net cost to his company as his time has been consumed - and of course the supplier of the accounting software will probably increase their charges because of the suppoed benefits.
We'll all have differing opinions but for me it often writes better emails and letters than myself, also it can save time so it's a win win for me (y)
 
A.I for sure can be dangerous in many different ways. Just as one example my brother brought in a coder to help with a new business start up, on the insistence of one of his investors the job was given to the investors son. The guy worked for him for about 7 months. As it turned out the guy didn't know what he was doing and was using A.I for everything. The project never got off the ground until my brother stepped in and the guy admitted that he didn't know how to fix all of the issues A.I had caused and he was sacked and replaced by someone that knew what they were doing. It is an A.I business so was a bit amusing.

For my own business we use A.I for both photo culling and editing via Aftershoot. We have been using it for about 18 months - 2 years or so and it saves a ridiculous amount of time if you are a wedding or events photographer, it can also do retouching as well. They are also about to launch their own online gallery system powered by A.I for delivering images to clients. This is my referral code if anyone wants to try it with a 10% discount. https://account.aftershoot.com/refe..._source=referral-page-app&utm_medium=copy-cta

We use all of the A.I options that are built into Photoshop and Lightroom.

We also use A.I for social media posts for content ideas, captions, scheduling etc.

We also use A.I for emails to clients, for some website content and marketing.

There is probably loads more we use it for in work that doesn't come to mind right now.

It has also replaced Google as a search engine for me and I use it for personal stuff many times every day.
 
Last edited:
A.I for sure can be dangerous in many different ways. Just as one example my brother brought in a coder to help with a new business start up, on the insistence of one of his investors the job was given to the investors son. The guy worked for him for about 7 months. As it turned out the guy didn't know what he was doing and was using A.I for everything. The project never got off the ground until my brother stepped in and the guy admitted that he didn't know how to fix all of the issues A.I had caused and he was sacked and replaced by someone that knew what they were doing. It is an A.I business so was a bit amusing.

For my own business we use A.I for both photo culling and editing via Aftershoot. We have been using it for about 18 months or so and it saves a ridiculous amount of time if you are a wedding or events photographer, it can also do retouching as well. They are also about to launch their own online gallery system powered by A.I for delivering images to clients. This is my referral code if anyone wants to try it with a 10% discount. https://account.aftershoot.com/refe..._source=referral-page-app&utm_medium=copy-cta

We use all of the A.I options that are built into Photoshop and Lightroom.

We also use A.I for social media posts for content ideas, captions, scheduling etc.

We also use A.I for emails to clients, for some website content and marketing.

There is probably loads more we use it for in work that doesn't come to mind right now.

It has also replaced Google as a search engine for me and I use it for personal stuff many times every day.
Yeah it's replaced google searches for me too.
 
I read an article the other other day regarding efforts to design a logo that identifies a product or service as "not generated by AI".

This IMO is the wrong course of action.

Every product or advert or news item that uses Ai should be forced by Law to have a warning label attached to it that It contains AI technology in its production. Ai is, for the gullible, far more likely to lead to societal harm than alcohol, drugs, tobacco, GM modified crops or mobile phones.

And, to the naysayers, if Ai is so great why are you not proud to boldly display its use by regulation?
 
In my day job as an IT Programme Manager we're encouraged to use AI, but to use it wisely and carefully.

An example of how I've been using it lately...we spin up a new project, I put the project kick off meeting in with all relevant stakeholders, the meeting is held on Teams and is transcribed.
During the call we'll talk about what the project is, what the key deliverables are, the main drivers behind it, discuss any key stakeholders, talk about any known assumptions, risks and dependencies, touch on the timeline and any known critical blockers.

I then take the transcription and feed it into Claude, I have a project setup in Claude so when I upload a transcription it produces a PowerPoint deck pre-formatted as per my instructions which will contains slides with the following:

  1. Project overview and objectives
  2. Key stakeholders and timelines
  3. Project assumptions
  4. Terms of reference for all project meetings (kick-off, steering, working group etc. (needed for Quality Assurance audit)
  5. RACI matrix
  6. Outstanding actions and flags
I'm not exaggerating when I say what it does in seconds using the meeting transcript could take a full morning or afternoon, possibly longer.

I also use it for emails, at times I'm not as succinct as I'd like to be, so I can give AI a summary of what I want to say and it will construct a succinct, professional email that sticks to the key points and in the tone intended.

It goes without saying that human verification is needed and I always read through and make any changes I think are needed, but when used correctly it's such an amazing tool.
 
Ai is, for the gullible, far more likely to lead to societal harm than alcohol, drugs, tobacco, GM modified crops or mobile phones.

And, to the naysayers, if Ai is so great why are you not proud to boldly display its use by regulation?
Agree with this 100%, people can very easily be fooled.
 
I'm sticking my feet firmly in the mud for now and trying to avoid it wherever possible, but my wife uses it at work for help with wording emails for team updates/course invites etc. She checks and tweaks everything that she sends out though so that it still sounds like her. Some of her colleagues have fallen into the trap of letting AI do all the work though, and she says that it's obvious when people haven't checked things properly, or just generated some text and sent it straight out.
This. if you can save 30 mins by using AI to generate it, use 5/10 of them to read through it and 'humanise' it if needed. You're still 'up' from a time point of view.
 
In my day job as an IT Programme Manager we're encouraged to use AI, but to use it wisely and carefully.

An example of how I've been using it lately...we spin up a new project, I put the project kick off meeting in with all relevant stakeholders, the meeting is held on Teams and is transcribed.
During the call we'll talk about what the project is, what the key deliverables are, the main drivers behind it, discuss any key stakeholders, talk about any known assumptions, risks and dependencies, touch on the timeline and any known critical blockers.

I then take the transcription and feed it into Claude, I have a project setup in Claude so when I upload a transcription it produces a PowerPoint deck pre-formatted as per my instructions which will contains slides with the following:

  1. Project overview and objectives
  2. Key stakeholders and timelines
  3. Project assumptions
  4. Terms of reference for all project meetings (kick-off, steering, working group etc. (needed for Quality Assurance audit)
  5. RACI matrix
  6. Outstanding actions and flags
I'm not exaggerating when I say what it does in seconds using the meeting transcript could take a full morning or afternoon, possibly longer.

I also use it for emails, at times I'm not as succinct as I'd like to be, so I can give AI a summary of what I want to say and it will construct a succinct, professional email that sticks to the key points and in the tone intended.

It goes without saying that human verification is needed and I always read through and make any changes I think are needed, but when used correctly it's such an amazing tool.
And therein lies the kick in the b*****ks,
Will you get the afternoon off that you saved by your inspirational use of the technology? If its your own company then fair play to you, if your employed then your afternoon will be crammed further until the point where AI agents are talking to themselves, managing all the aspects of the process you list and you will slowly realise that you are no longer required.

AI is mostly Turkeys voting for Christmas
 
Last edited:
And therein lies the kick in the b*****ks,
Will you get the afternoon off that you saved by your inspirational use of the technology? If its your own company then fair play to you, if your employed then your afternoon will be crammed further until the point where AI agents are talking to themselves, managing all the aspects of the process you list and you will slowly realise that you are no longer required.

AI is mostly Turkeys voting for Christmas
Not at all, it just frees up my time to do other things that I need to do, more time to sit and scratch my arse :ROFLMAO:

I still have to drive projects and make sure tasks happen, risks are analysed / mitigated and a whole range of other tasks, AI just helps me with some of the more mundane activities, so I don't have to sit on a Teams call and furiously try to scribble notes to capture actions while also trying to follow the conversation, that's a huge win for me.
 
Not at all, it just frees up my time to do other things that I need to do, more time to sit and scratch my arse :ROFLMAO:

I still have to drive projects and make sure tasks happen, risks are analysed / mitigated and a whole range of other tasks, AI just helps me with some of the more mundane activities, so I don't have to sit on a Teams call and furiously try to scribble notes to capture actions while also trying to follow the conversation, that's a huge win for me.

For now,

What will you be doing in a Year? Three years ? Five years?
 
What will you be doing in a Year? Three years ? Five years?
Tramp on bench.jpg

Something like this, if the greedy people have their way?
 
Hopefully still breathing, but there is no way AI can replicate everything we do in our organisation.
People said the same thing about the internet, I was around then and remember it clearly.

So do I and looking back, it was underutilized, pathetic and feeble, Which is exactly what we will be saying in the future looking back at this, the beginning of AI,

Lets hope that Ai solves the quantum computing problems so that its unrestricted by the limitations of our current technology and the nuclear fusion problems so that it has unlimited power (sarcasm).

It will b******t some clown into building it.
 
It will b******t some clown into building it.
So far as I'm concerned, the real danger of artificial "intelligence" programs, lies in their being black boxes without, it appears, there being any way to debug them.

Worse still, the end user seems to have no way to find out why an AI device made any decision, which is just plain stupid, in my book. Nobody should be allowed to build machines that no-one can check and debug. This is where the law makers can and should take action, before things go wrong. The UK Post Office fiasco shows what can go wrong, when the criminally minded hide the details of machine operations from the customer (in this case, the postmasters and postmistresses). In that case, there was a way to track the mistakes, although the perpetrators spent a lot of time denying it. So far as I can tell, there will be no such path to the truth, with AI systems.
 
We'll all have differing opinions but for me it often writes better emails and letters than myself, also it can save time so it's a win win for me (y)

The problem is that the advantages neutralise themselves. Bit like academic grade inflation.

Give everybody and an A and the A becomes worthless. But worse. If everybody is paying for a subscription service to get that 'A' standard that is then of no value then it just adds a cost with no net benefit.
 
Yeah it's replaced google searches for me too.

It's noticeable that searching for something where there might be useful or subtle results has become harder.

And not all summarised results are accurate or complete.
 
The problem is that the advantages neutralise themselves. Bit like academic grade inflation.

Give everybody and an A and the A becomes worthless. But worse. If everybody is paying for a subscription service to get that 'A' standard that is then of no value then it just adds a cost with no net benefit.
As @snerkler said, I will compose an email, get my points across, sometimes I may waffle a bit :) so I can run it through AI and ask it to re-compose it and make it shorter, or any other changes I'd like to change, but I'm feeding in the information.

Same when I use it for my project meetings, I'm feeding in the data, which is the transcription from the meeting, AI isn't making it up, it just reads it, understands it and frames it in a way that I've pre-defined.

Like everything it can be a good tool, if used sensibly and always with caution.
 
The problem is that the advantages neutralise themselves. Bit like academic grade inflation.

Give everybody and an A and the A becomes worthless. But worse. If everybody is paying for a subscription service to get that 'A' standard that is then of no value then it just adds a cost with no net benefit.
I can see that with regards to academics, but that's different to composing emails and letters. These aren't graded, they're not compared, I just want it to sound a particular way and sometimes I cna't think of the right words or phrasing. I don't see how a well worded email is worthless regardless of how many well worded emails there's been before or after :thinking:
It's noticeable that searching for something where there might be useful or subtle results has become harder.

And not all summarised results are accurate or complete.
That's true of google too, very little on the internet should be taken as gospel and due diligence it required whatever you use (y)
 
From my point of view, AI is just another tool, albeit a very powerful one. I’ve worked with automation and analytics for years, so I’m quite comfortable with it, and for me the same rule applies as with any automation, i.e. test it, review and amend as needed, and above all own the output. For photography/video business, I find the strongest use cases up to now are on the admin side proposals, contracts, planning, coding small automations, and there is potential for culling/selection. I also like that machines for photo/video work are good for running local models where privacy matters. Used well, it saves time; used lazily, it just produces faster mistakes.
 
From my point of view, AI is just another tool, albeit a very powerful one. I’ve worked with automation and analytics for years, so I’m quite comfortable with it, and for me the same rule applies as with any automation, i.e. test it, review and amend as needed, and above all own the output. For photography/video business, I find the strongest use cases up to now are on the admin side proposals, contracts, planning, coding small automations, and there is potential for culling/selection. I also like that machines for photo/video work are good for running local models where privacy matters. Used well, it saves time; used lazily, it just produces faster mistakes.
Agreed, you just worded it much better than me....did you use AI? :ROFLMAO:
 
PS I came across an incredible use case for AI yesterday to fix tricky green screen problems. This has always been a PITA when filming people with fine hair, lots of motion blur or semitransparent situations. A small team has created a solution that they have made freely available on github. As I understand it, they automated the production of thousands of example tricky frames to train the model and the results are incredible. For me, this is a great example.
 
Back
Top