• Skip to main content
  • Skip to primary sidebar
  • Home
  • Articles
  • Blog
  • About
  • Contribute
  • Contact

Sceptical Scot

Asking Questions. Seeking Answers.

You are here: Home / Blog / AI, Creativity and Humanity (3): Responsibility

AI, Creativity and Humanity (3): Responsibility

October 19, 2025 by Dougal Perman Leave a Comment

In the third and final part of his series, Dougal Perman argues the case for combining creativity and culture with being fluent in AI – and staying human.

When it comes to creativity, AI may be a great enabler but it can’t read your mind (yet) and definitely won’t do everything for you.

And nor should it. Like everything, you get out what you put in. Some people disregard AI because they think it produces poor results after they try something once without putting in any effort.

AI is a technology. We can use it as a tool to help us do more, better, faster, cheaper work. But we need to master the use of that tool to get the best results. An MIT study shows that ChatGPT can boost productivity for writing. However, a small but much quoted MIT Media Lab study found that people who use ChatGPT to write their assignments had a low sense of ownership of, and inability to quote from, the work. You don’t say? Getting ChatGPT to write your essay for you is the modern equivalent of paying another student to write your essay for you.

You wouldn’t walk into a library and expect the books to jump off the shelves and tell you what you need to know. You have to work with LLMs to get the best out of them. Valuable data analysis and strategic insights may come relatively quickly but the iterative prompt crafting, refining, revising and improving (i.e. “prompt engineering”) takes time to learn and implement. I can get game-changing results from AI tools, but it can take hours.

Talking to LLMs

Natural language processing is the method by which we communicate with LLMs. Working on structuring research to help a funding application we were working on, a composer and performance artist friend became excited about the way I was interacting with ChatGPT.

“It’s like writing code in English. You’re programming with normal language!”

I combine strict instructions and clearly defined parameters with my desired outputs, for my intended audience, and wrap it in everyday language. That, to me, is the beauty of using LLMs. Being able to talk to computer programmes in everyday language is incredible. Especially if you understand that all the incredible things LLMs can do (coding, writing, data-mining, creating, etc.) are the result of token prediction.

Innovating With AI founder Rob Howard (whose AI Consultancy course I’m doing) describes next token prediction like a “plinko board” game (a bit like The Price Is Right). No matter how sophisticated the software is, it is just software. LLMs aren’t self-aware, or sentient. They are not human. They’re not Johnny 5 (or the Hybrids in Alien Earth).

Much of the online chatter about AI (across Reddit, social media, blogs, traditional media, etc.) is overhyped — the launch of ChatGPT 5 definitely was. It’s very good, but it’s not the iconoclastic wonder it was portrayed as. What’s more concerning, however, is the backlash OpenAI got for discontinuing ChatGPT 4o. Read reactions on Reddit like “chatGPT is my only friend”. At best it’s funny, but mostly it’s pretty worrying. The backlash was so vociferous that OpenAI reinstated 4o as a legacy model option, perhaps unwisely. AI should not be used as your sole therapist. It can be useful, productive or even just fun to converse with an LLM, but it is not a person. We need to teach people that AI is just technology. Creativity could be a way to do that.

Communicating through poetry

In 2006 I was invited to join a high level future of education workshop run by the senior team of a department in Scottish Enterprise that went on to be Skills Development Scotland. One of the speakers said “in the future computers will be voice activated so instead of teaching school children (indeed, all of us) to code, we should teach them poetry so they can communicate properly.” How right he was. Many times on stage, in groups and one-to-one I have advocated for project-based learning that focuses on teaching everything through the medium of culture (in which I include music, storytelling, drama, craft, art and design, dance, sport, cooking, gardening and so on). Culture is how we understand and interact with the world around us. Embracing it is more important now than ever.

My son is in Primary 7. He wants his school to teach them how to use AI and how to become a YouTuber. I support both ambitions. Through project based learning we can teach young people how to make videos to tell their stories, using AI as a tool in the process. (I’m up for helping schools do this, give me a call if you want help embracing AI, Education Scotland.)

AI isn’t going anywhere. You interact with it many times a day whether you know it or not. You don’t need to fear it and I encourage you to embrace it. But you should also challenge it, and those who are developing it. Keep an open mind, balance optimism with scepticism and use creativity to get the most out of AI.

Culture is at the heart of humanity. When Johnny 5 came to life, he learned human values through immersing himself in culture. By keeping our creativity in the foreground, we can forge useful relationships with LLMs (and AI robots when they roll out). And by championing our culture, we will always preserve our humanity.

Further info/reading:

Films/Culture

  • Short Circuit (1986) — IMDb: https://www.imdb.com/title/tt0091949/

Concepts

  • “AI slop” — Wikipedia: https://en.wikipedia.org/wiki/AI_slop

Breakthroughs/Research tools

  • AlphaFold Protein Structure Database: https://alphafold.ebi.ac.uk/

Your infographic

  • Where AI Adds Real Value in Creative Work (Gamma): https://gamma.app/docs/Where-AI-Adds-Real-Value-in-Creative-Work-4h7zw66y4po0h8p

Energy, water, system-level impact

  • IEA — Energy and AI: Energy demand from AI (chapter): https://www.iea.org/reports/energy-and-ai/energy-demand-from-ai IEA
  • FT — Camilla Palladino, How AI might save more energy than it soaks up: https://www.ft.com/content/69d6a641-971c-4fe8-a58d-0f3377a6bccb
  • OpenAI / Sam Altman — The Gentle Singularity (per-query energy & water): https://blog.samaltman.com/the-gentle-singularity Sam Altman
  • Epoch AI — How much energy does a ChatGPT query consume? (≈0.3 Wh): https://epoch.ai/gradient-updates/how-much-energy-does-chatgpt-use (alt explainer: https://www.linkedin.com/posts/epochai_how-much-energy-does-a-chatgpt-query-consume-activity-7300899999268167680-JRTh) LinkedIn

Employment/Adoption

  • Klarna press release — AI assistant handles two-thirds of customer service chats in its first month: https://www.klarna.com/international/press/klarna-ai-assistant-handles-two-thirds-of-customer-service-chats-in-its-first-month/ Klarna
  • OpenAI case study — Klarna’s AI assistant does the work of 700 FTEs: https://openai.com/index/klarna/ OpenAI
  • Business Insider — Klarna reassigns workers to customer support after AI push went too far:https://www.businessinsider.com/klarna-reassigns-workers-to-customer-support-after-ai-quality-concerns-2025-9Business Insider
  • FinTech Weekly — Klarna resumes hiring CS agents after AI pivot:https://magazine.fintechweekly.com/magazine/articles/klarna-hires-customer-service-after-ai-pivot FinTech Magazine Articles
  • ITPro — Australia’s biggest bank cut staff for AI, then backtracked: https://www.itpro.com/business/business-strategy/australias-biggest-bank-cut-staff-for-ai-then-it-backtracked-and-its-one-of-many-scrapping-plans-for-automated-customer-support-teams IT Pro

Productivity & learning

  • MIT — Study finds ChatGPT boosts worker productivity for writing tasks: https://news.mit.edu/2023/study-finds-chatgpt-boosts-worker-productivity-writing-0714 MIT News
  • TIME — ChatGPT may be eroding critical thinking, says new MIT Media Lab study (early results):https://time.com/7295195/ai-chatgpt-google-learning-school/ TIME

NLP/LLM explainers

  • ai — Natural Language Processing (overview): https://www.deeplearning.ai/resources/natural-language-processing/
  • NVIDIA Blog — AI Tokens Explained (next-token prediction): https://blogs.nvidia.com/blog/ai-tokens-explained/

Quotations (timestamped video)

  • Mo Gawdat on Diary of a CEO — time-stamped quote: https://youtu.be/S9a1nLw70p0?t=4291
  • Dr Joan Palmiter Bajorek — Your AI Roadmap episode (trades that are hard to automate), time-stamped:https://www.youtube.com/watch?v=DM-xpISOvNc&t=588s

Model availability/user backlash

  • Ars Technica — OpenAI brings back GPT-4o after user revolt (Aug 2025): https://arstechnica.com/information-technology/2025/08/openai-brings-back-gpt-4o-after-user-revolt/ Ars Technica

(Optional) Reddit context

  • “is it just me but chatgpt is honestly my only true friend”: https://www.reddit.com/r/ChatGPT/comments/1aifildReddit

 

Filed Under: Blog, Culture, Featured, Media Tagged With: AI, arts, Scotland

About Dougal Perman

Dougal Perman is a creative industries consultant, director of Inner Ear and former chair of the Scottish Music Industry Association.

Reader Interactions

Leave a Reply Cancel reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Primary Sidebar

About Sceptical Scot

Welcome to Sceptical Scot, Scotland’s premier non-tribal forum for passionate, informed debate. Sceptical Scot is for all who care about Scotland’s future, regardless of how they vote: for party, independence or union, EU or Brexit. We aim to provide an arena that is both broader and deeper than current online/print offers with a rich diet of well-researched, polemical, thought-provoking writing. Read more » about About Sceptical Scot

What’s new on Sceptical Scot

  • AI, Creativity and Humanity (3): Responsibility October 19, 2025
  • AI, creativity and humanity (2): Obstacles October 17, 2025
  • AI, creativity and humanity (1) October 16, 2025
  • Storm Amy and 75K escaped Scottish salmon October 15, 2025
  • Two worlds collide or stories Geoff Palmer told October 15, 2025
  • Israel’s biggest security threat: Netanyahu October 7, 2025
  • Middle East transformed in two years of war October 7, 2025
  • Build, baby, build October 5, 2025
  • Does Starmer have a vision? October 4, 2025
  • Political change across Britain and Ireland (2) September 24, 2025

The Sceptical Newsletter

The Sceptical Scot cartoon

Categories

  • anti-fascism (12)
  • Articles (722)
  • Blog (677)
  • Books & Poetry (27)
  • Brexit (231)
  • climate crisis (9)
  • climate crisis (55)
  • Covid19 (67)
  • Criminal justice (19)
  • Culture (341)
  • Devo20 (1)
  • Economics (202)
  • Economy (178)
  • Education (84)
  • Elections (242)
  • Energy (12)
  • Environment (105)
  • European Union (294)
  • Featured (44)
  • Federalism (23)
  • federalism (15)
  • Health (71)
  • History (97)
  • Housing (29)
  • Humour (11)
  • identity (32)
  • Independence (322)
  • Inequality (88)
  • International (81)
  • Ireland (15)
  • Ireland (8)
  • Local government (97)
  • Longer reads (79)
  • Media (18)
  • Podcast (4)
  • Poetry (73)
  • Policy (309)
  • Politics (470)
  • Polls and quizzes (2)
  • protest song (1)
  • Reviews (26)
  • Social democracy (93)
  • Trump (25)
  • UK (423)
  • Uncategorized (15)

Sceptical Scot elsewhere

Facebook
Twitter

About Sceptical Scot

Since 2014 Sceptical Scot has offered a non-tribal forum for passionate, informed debate for all who care about Scotland’s future

Copyright © 2025 · Magazine Pro on Genesis Framework · WordPress · Log in

 

Loading Comments...