In fact the AI generated posts are better than the old school bot produced rubbish, which are simply ads with no relation to H&S. At least these touch on our subjects. The problem is they are not always obviously AI generated. You can spend time thinking of a response only to realise that it is AI generated after they include some Americanisms etc. some might actually be honest attempts by people whose first language is not English trying to get information.

As the amount of AI content increase there will be a tendency for the thread to be initiated by AI and then all the responses to be generated by AI  which rather defeats the object of a public forum.

Originally Posted by: peter gotch

Hi HSSnail

Have you emailed IOSH about this concern or would you consider doing so?

If the contact details in the Forum Rules are up to date then the email address is

socialmedia@iosh.comThanks Peter, no have not got round to it, it just ticked my silly bone that AI would actually tell me the problems with itself.

Shane, are you from Safety First UK a private consultancy? This look suspiciously like your trying to hide an advert? This is against forum rules.

Shane

I took the precaution of printing off the entirety of this thread yesterday, so I have been able to paste both versions of your initial posting into a Word document to compare them and work out what amendments you have made.

I think that the changes are such that you are probably no longer in breach of the Forum Rules.

I suppose reducing the length of a single paragraph must make it marginally easier to digest what you wrote but only marginally and I hope that this "demonstration" app is better at writing in something akin to Plain English.

However, I can't see how what you have posted has any relevance whatsoever to the thread that HSSnail started.

So, I wonder what purpose you thought and perhaps still think there was in replying to what HSSnail had written?

I would note that as someone whose very close involvement in the practical implementation of what started off life with the acronym CONDAM before that was ditched for CDM by the time the first iteration of the Regulations came into force in Great Britain in 1995, it seems to me that this app is highly unlikely to do anyting that hasn't been done on numerous occasions over the last 25 years 

Perhaps worse, you indicate that those implementing CDM in the UK and the Construction Regulations in South Africa might benefit from the app referring them to UK historic accident data. 

First that concerns me as not once in your posting do you mention the word "health" which suggests that this may not be an issue of concern for you or the app, when the available authoritative data from e.g. HSE in Great Britain and the ILO tells us that the toll in harm is skewed heavily to adverse impacts of occupational health risks and NOT to accidents.

Then again, I am not sure how people would use historic accident data from the UK in the absence of appropriate context that would enable prospective risks to be quantified during project development and then execution stages.

So, as a simple example, decades ago, roughly 10 people a year would die in the UK as a result of falling whilst doing "roofwork". However, never were the statistics broken down to indicate just how each of those 10 falls a year would be occurring, so as example one MIGHT have been from a scaffold complete with all expected precautions, another MIGHT have been from a sloping roof with no edge protection. [I have invesigated fatal accidents involving each such scenario, and you might think it somewhat odd that I wrote a prosecution report for the first but NOT for the second].

....oh and eight other roofwork fatal accidents. All from falls. Very little other than that in common.

For each of those deaths in fatal accidents probably 10 people whose tasks involved "roofwork" would have died or will die prematurely as a result of exposure to occupational health risks.

Those on construction project might arguably learn if we could confidently calculate the Individual Risk of Death per annum (IR - cf HSE's R2P2) of those doing roofwork but we can't for the simple reason we don't have a reliable number for the population at risk - we DO know roughly how many people are classified as "roofers" but many people other than "roofers" do roofwork.

HSE has gone to great lengths over the years to try to get through to CDM duty holders that those in the project development stage should not be telling competent Contractors how to suck eggs, so if there is roofwork to be done, they shouldn't be flagging this up and perhaps referencing the guidance in HSG33, unless there is some reason why the risks are going to be unusual or difficult to manage.

Is this app going to guide a Client or Designer engaged on a project whose product will be a new building, as to why sometimes designing in permanent edge protection for the roof might be a good idea, but equally why it might NOT be?

OR is this app just another means of ensuring that construction projects are buried under FAR too much H&S paperwork that is usually counterproductive as the important information tends to be obscured by lots of bureaucratic waffle?

How often have you been involved in a substantial CDM project from inception to completion and working out how many CDM Health and Safety Files will be needed, who needs them and what should be contained in each? If you think that CDM says that there should only be one such File I invite you to read through the Interpretation Act 1978 and may be guidance from the likes of CIRIA.

Let's suppose that your app user has a road to build. Will the app help them decide on what precautions to apply in relation to each of the culverts under the road? Would it be a good idea to provide a screen at the intake to stop falling trees and other debris from blocking flow through the culvert?

Of course, if you want to stick to buildings, then I could ask exactly the same question in terms of potential obstructions to the flow to the downpipe from the roof.

Edited by user 06 November 2025 19:43:23(UTC)  | Reason: An errant + got in

Shane, I quote:

"however if one includes we will wash down the scaffolding with carbon Tetra chloride, the app is built in such a way that HSE.ex stats will be consulted to see if any Occupational health conditions have been caused,"

Can you give an indication of what statistics you might consult?

Apparently you are an Occupational Hygienist Environmental Scientist and Safety Professional. If that is the case, then surely you would know that in the main the statistics that are collected (and particularly those that look into causation) are for the adverse impacts that present quickly.

So, if the carbon tetra chloride you refer to being used to wash down scaffolding results in e.g. someone being overcome possibly as a result of confined space working, it MIGHT end up in the stats. In contrast, if someone gets respiratory conditions or dementia many years later it is very unlikely that you will find it in the statistics. Ditto if people end up with adverse impacts to the skin even in the short term.

I have worked on upwards of 1000 CDM projects with values ranging from £5000 (very low because the hard work on site was to be done by volunteers) to £3.5bn, so of course I have seen the method statements that are not fit for purpose and sometimes relating to different work, or even a different project. Those managing the project don't need some app to identify the glaring deficiencies. They just need a reasonably competent project team (and I don't mean the OSH professionals, even if often the Safety Bod may get the dross dumped on them by line managers who want to offload their responsibilities).

I have also seen plenty of examples of either proprietary systems or templated solutions marketed by so called health and safety consultancies that attempt to deal with all the various deliverables that are needed on a CDM project. Invariably they have been suboptimal. Partly as "construction" projects have so many variables to take into account. 

You have worked in O&G. Did any of those you worked with produce monthly statistics predicting how many of the workforce would later suffer from occupational ill health? Ditto the public and environment nearby? 

This afternoon I have looked at a multnational's 2025 Sustainability Report. Absolutely no surprises in the content. The Key Performance Indicator for H,S and E that was apparently of greatest importance was the LTIFR and that was going down nicely. Not ONE WORD on how many of the workforce were being exposed to occupatonal health risks, not about the actual risks to people and the environment beyond the gates of the manufacturing sites. So, even if we ignore the health risks to the public, nothing to give the reader any comfort that the company is managing its supply chain when it comes to the road traffic risks of vehicles taking things to and from the sites.

On many construction projects it is MUCH more likely that someone will die from an accident OFF site than ON.