This is something our Marketing Manager created as little humorous piece written entirely for fun!
The Secret Diary of Gary Clencher, I.T. Project Director, Aged 51 and ¾ is a fictional diary series following one very real type of person through one very recognisable type of project. Inspired by Sue Townsend. Written entirely for fun.
If you’ve ever sat through a steering group that should have been an email…
If you’ve ever nodded along to “human-centred AI orchestration”…
If at least one person in your organisation is pretending this isn’t happening…
This one’s for you.
A Note Before We Begin
This is entirely Helen’s fault.
It started, as many questionable ideas do, on a Tuesday evening somewhere around the second glass of Primitivo, when your Marketing Manager found herself wondering what it would actually feel like to be a middle-aged IT Project Director trying to deliver an AI transformation programme while the organisation slowly rearranges itself around him.
The answer, it turns out, is Gary Clencher.
The Secret Diary of Gary Clencher, I.T. Project Director, Aged 51 and ¾ is a work of pure fiction – a comedic love letter to everyone who has ever sat through a steering group that should have been an email, received a change request with forty-seven tracked amendments, or nodded along in a meeting about “human-centred AI orchestration” without the faintest idea what was being orchestrated or for whose benefit.
Gary, Sandra, Barry, Karen, Priya, Trevor, Darren and the rest are entirely made up. Any resemblance to actual project directors, PMO analysts, or middleware is coincidental and, frankly, says more about your organisation than it does about ours.
The whole thing was inspired by the incomparable Sue Townsend, whose Adrian Mole taught a generation that the gap between how we see ourselves and how the world sees us is the funniest place there is. Gary Clencher operates firmly in that tradition – except instead of spots and unrequited love, he has scope creep and a middleware issue that Darren refuses to acknowledge.
Read it as it was written: with a glass of something, and absolutely no expectation of professional development.
Helen Marketing Manager, Glass Half Full
The Secret Diary of Gary Clencher, I.T. Project Director – Aged 51 and ¾
Navigating enterprise systems, impossible deadlines, and the terrifying spectre of artificial intelligence. A completely confidential document.
Do not share with Patricia from Finance.
Week One
In which Gary discovers AI, offends Karen, and emails the wrong budget to Patricia.
Monday
Have decided to embrace AI. Barry from the Change Management team said it will “transform our ways of working.” Barry once thought a pivot table was a piece of gym equipment. I am nonetheless keeping an open mind.
Spent forty-five minutes trying to log into the AI platform we purchased in Q3. IT helpdesk said my account “exists in principle.” This is not the reassurance I needed.
Am 51 and three-quarters. Have successfully delivered seventeen major IT programmes. Have once fixed a photocopier. I will not be defeated by a chatbot.
Tuesday
Got into the AI platform. Asked it to summarise the project risk register. It produced a seventeen-page document of extraordinary confidence and moderate accuracy. Like a consultant, but faster and without the expense account.
Note to self: “Hallucination” in AI does not mean the system is unwell. Have googled this. Still not entirely reassured.
Showed it to Karen from the PMO. She said it had invented two risks she’d never heard of and omitted the one about the server room flooding. I pointed out this was arguably an improvement on Karen’s own register, which she did not take well.
Project board meeting at 3pm. Have told no one about the AI risk register. Am sitting on a secret like a hen on a digital egg.
Wednesday
Big day. Presented the AI tool to the wider programme team. Described it as a “force multiplier for intelligent delivery.” This phrase came from the AI itself, which I feel is somewhat circular.
Darren from Infrastructure asked if it could fix the legacy middleware issue that has plagued us since 2019. I asked the AI. It said “this is a complex and multifaceted challenge requiring stakeholder alignment.” Darren said that was exactly what I always say and walked out.
Tried to get AI to write the weekly steering committee update. It did so in under eight seconds. The update was grammatically perfect and utterly devoid of the defensive ambiguity I have spent years perfecting. Had to edit heavily. Added three uses of “ongoing,” two of “in flight,” and one “we are where we are.”
Thursday
Have been asked to present our “AI adoption strategy” at next month’s Digital Futures conference. This is extremely ironic as our current AI adoption strategy is me, alone, at my desk, asking a chatbot to explain what “RAG architecture” means and nodding slowly.
RAG stands for Retrieval-Augmented Generation. This does not mean what I hoped it meant. Nobody in the project environment appears to know what it means either, but everyone is saying it with great authority.
Called a team meeting to discuss “AI integration across the programme.” There were twelve people. Ten laptops. And a collective understanding of AI that could be generously described as folkloric.
Have drafted a roadmap. The AI helped. It added a phase called “Value Realisation” which sounds excellent and means nothing. We are in business.
Friday
Quiet day. Used the AI to rephrase seventeen emails. They were all better. This is troubling.
The AI has suggested we automate the monthly status report using a pipeline that “ingests project artefacts and surfaces actionable insights.” I have forwarded this to the Head of Portfolio, describing it as my idea. This is wrong of me. I am not a good person.
Spent the afternoon reading about “prompt engineering.” It is apparently a skill. I have been doing it wrong for months. You must be specific. You must give context. This is, I note, exactly what I tell my team about writing project briefs, and they also ignore it entirely.
End of week one. Have learned the following: AI is very good at writing things. AI is moderately good at numbers, if supervised. AI does not understand Patricia from Finance. In this, AI and I are united.
Week Two
In which AI meets the board, the budget catastrophe unfolds, and Trevor arrives.
Monday
Have been asked to explain machine learning to the board. Have eleven days to prepare. This is both too long and not nearly long enough.
Spent the morning asking the AI to explain machine learning in simple terms. It gave me a very clear explanation involving dogs and cats. Asked it to make it more professional. It removed the dogs and cats and added “pattern recognition” seventeen times. Felt like a loss.
Machine learning is, as far as I can tell, the process of showing a computer thousands of examples until it stops being wrong. This is also how I trained Dave on the helpdesk. The analogy is uncomfortably apt.
My wife Sandra asked how work was going. I said I was “leading the organisation’s AI transformation agenda.” She asked if I’d remembered to put the bins out. I had not. Even AI cannot help me with this.
Tuesday — The Budget Incident
A grave day. Will record events faithfully for posterity and possible use in a tribunal.
Asked the AI to assist with reforecasting the programme budget. Gave it the spreadsheet. Gave it the context. Gave it what I thought were clear instructions. The AI reforecast the budget with tremendous efficiency and quiet confidence. It moved £340,000 from the contingency reserve into “Innovation and Experimentation.” I did not notice this. I sent it to the Finance Director.
The Finance Director’s name is Patricia. Patricia has worked in finance for twenty-six years. Patricia does not experiment.
Patricia’s email response contained four questions, two of which were rhetorical, and one sentence that ended with the words “…if this is indeed a serious document.” Have read this eleven times. It does not improve.
Replied to Patricia saying the reforecast was “a working draft intended to provoke strategic dialogue.” This is not true. It was meant to be the final version. I had already put it in the board pack.
Removed it from the board pack at 11:47pm. This is why I cannot sleep. Am 51 and three-quarters. My contingency reserve is now in innovation. This is a metaphor for something.
Wednesday
Recovery day. Told the team the budget “remains under active review.” Nobody asked further. This is either trust or indifference. After seventeen years in IT delivery, I cannot tell the difference.
Returned to the board presentation. The AI has now produced a twelve-slide deck explaining machine learning. Slide four contains a diagram that is technically accurate and completely incomprehensible. It looks like a plumbing schematic for a building that hasn’t been built yet.
Showed it to Nigel, who is our Enterprise Architect and therefore professionally obliged to find it interesting. Nigel said it was “a reasonable representation of a neural network.” I asked if the board would understand it. Nigel said “probably not, but it’ll make them feel like they should.” Kept the slide.
Have added a note at the bottom of slide four reading “simplified for clarity.” This is a lie. Nothing has been simplified. I have merely made the arrows slightly thicker.
Thursday — The Board Presentation
It is done. I am alive. This is the best that can be said.
Presented “Harnessing Artificial Intelligence Across the Programme Landscape” to the full board. Fourteen people. Twelve laptops. One man called Trevor who I have never seen before and who was introduced only as “from the centre.” Trevor wrote nothing down. This frightened me.
The dogs-and-cats explanation went well. Sir Michael laughed and then looked embarrassed about laughing, which I took as a good sign. The neural network diagram caused a silence of approximately eight seconds, which in a boardroom is geological.
The Deputy Chief Executive asked if AI would “replace the project team.” I said, “AI augments human capability rather than substituting professional judgement.” The AI wrote this line for me. There is a poetry to this that I choose not to examine.
Trevor from the centre asked one question at the end: “What’s your data governance framework for AI-generated outputs?” I said we were “actively developing a robust and iterative framework.” Trevor wrote this down. This was the only thing Trevor wrote down. I will be thinking about Trevor for some time.
Went home. Had a biscuit. Had three biscuits. Stared at the wall. Am 51 and three-quarters. Have survived.
Friday
Received two emails saying the presentation was “really thought-provoking.” In my experience “thought-provoking” means they don’t know what they think yet. This is fine. Neither do I.
Barry from Change Management has announced he is now our “AI Champion.” He has changed his email signature. It includes a small robot emoji. I said nothing. One must choose one’s battles.
Karen from the PMO has started using the AI tool. She asked it to rewrite the project initiation document. The AI produced something clear, well-structured, and free of passive voice. Karen said it “didn’t sound like us.” She is correct. This is the problem and also possibly the solution.
End of week two. The board presentation is complete. Trevor has been encountered. The budget incident is unresolved. Barry has one robot emoji. The organisation’s relationship with AI can be described as “ongoing.” Am 51 and three-quarters.
A Final Word (For Now)
I’m writing this from the departure lounge, metaphorically speaking — hand luggage packed, out-of-office on, and a novel set in Space already open on my Kindle.
Gary, however, is not going on holiday. Gary never goes on holiday. There will be something in his inbox when he gets back from his, and it will almost certainly be from Karen, and it will have tracked changes.
Whether Gary’s story continues beyond Week 2 depends entirely on you. If you’ve enjoyed his particular brand of suffering — the bilateral meetings with Trevor, Barry’s domain name, Darren’s middleware developing what we can only describe as a personality — then do let me know. Forward it to someone. Share it with whoever in your organisation would recognise every single character without needing a cast list.
If it lands well, I’ll pour another glass and find out what happens next.