A few days from now, I’m starting the ninth cohort of AI for Personal Productivity, a 12 hour training programme for the Irish Research Development Group. This content below was published from my Obsidian vault with one click. I’m sharing the content here to show the end result of my unedited workflow from numbered notes directly to my microblog.

I’m learning how Markdown syntax translates from Obsidian to HTML. There are some accidental text highlights that come across when I publish.

Helicopter View

Info from IRDG Innovation Skillnet

2. Your Facilitators

Larry Maguire

• Organisational Psychologist, Lecturer in  psychology, leadership coach, workshop facilitator.
• w: humanperformance.ie

Bernie Goldbach

• Public Service Technologist. Co-founder Iteration Ltd.
• w: InsideVidew.ie

3. Today’s Session

0. Background information.

DEMO:  Four AIs have developed content for today.

BREAKOUT 1:  Burning Questions

1. Digital Transformation and AI Workflows
2. The Four-P Framework For Successful Digital Transformation

  DEMO:  Claude writes content.

3. GCAO Framework for prompting AI

  BREAKOUT 2: Use AI to ask about this programme

4. Perspectives and Recap

4. Learning Outcomes

Understand how to implement AI into personal and professional workflows to enhance productivity.

Learn how to leverage generative AI through prompts that generate relevant text and images.

Know the purpose and use cases of several tools that are enhanced with AI functionality.

Appreciate the human factors and workplace impacts associated with AI and digital transformation.

Know the fundamental requirements for improved personal digital effectiveness.

5. Background: Pro Tips

• Learn to do more than generate content. Do it sustainably.
• Process-based outcomes are foundations of digital transformation.
• Use AI for personal productivity.
• Appreciate EU AI Act.

7. Background information

This training incorporates business intelligence since 2010 in the fields of work psychology and machine learning. We have been actively using workflow automation tools since 2020.

More than half of this course has changed since its first running in January 2024.

This training incorporates business intelligence since 2010 in the fields of work psychology and machine learning. We have been actively using machine learning tools since 2022.

9. Breakout 1: Why are you here?

Let’s have a small group discussion. We’d like to know…

• What digital tools are you using? -What’s your burning questions around digital transformation?
• What concerns you about using AI?

10. Previous Concerns

• Breach of GDPR and confidentiality.
• Copilot mostly used. Company concerned about privacy, hallucinations, keeping up with new developments.
• AI Meeting Assistants
• Your phones could be listening to this discussion too.

11. Questions We Will Answer

• What is Generative AI?
• How does it relate to personal digital transformation?
• How do digital transformation and AI affect people?
• What skills, attitudes & behaviours do we need to develop?
• What tools and apps can help our workflows?

12. DEMO:  We asked Copilot “What tools?”

13. How AI models evolved

• 2018 showed how to pre-train an AI
• 2020 pre-trained models could respond to prompts without any new training
• 2022 Reinforcement Learning from Human Feedback
• 2023 Multimodal AI and Slow Thinking (started IRDGAI)
• 2025 Agentic AI (agents and chatbots)
• 2026 System 2 Thinking (explained in this course)

14. Generative Pre-Trained Transformer

15. About context

• Context represents the short term memory used by an AI. If you exceed ‘context window’ limits, the AI starts to forget.
• As the video shows, an AI uses ‘context’ to determine meaning.t any new training
• You will learn to help an AI refine its context by using efficient prompts.

17. About ChatGPT

• You can talk to ChatGPT with your natural language.
• Multimodal input and output - can interpret images, create diagrams, and process code.
• Pro can retain info across sessions via customisable memory - greater personal relevance over time.

18. Time to reach 100m users

ChatGPT currently has over 200 million weekly active users, according to Axios. This represents a doubling of its user base since November 2023, when it had approximately 100 million weekly active users.

 OpenAI says ChatGPT usage has doubled in the last year (axios.com)

19. Public opinion about AI is waning

• American public opinion toward artificial intelligence is now less supportive than it is for Immigration and Customs Enforcement (ICE), a government agency that has historically faced significant public criticism (Cybernews, 2026; Gizmodo, 2026).
• Only 26% of registered voters reported feeling positively about AI, while 46% held negative views (Cybernews, 2026).

20. About Microsoft Copilot

• Integrates with the Microsoft 365 ecosystem
• Enterprise-Grade Security and Compliance
• Advanced Personalisation and Contextual Understanding

21. About Google Gemini

• Gemini has the biggest context window.
• If you have Gmail or you routinely use Google Search, you should add Gemini to your tool box.

22. Claude AI

• Claude uses a unique safety technique called Constitutional AI.
• Its models have extremely large context windows and that means it can process large documents or complex codebases in one go.
• Writers cite its human-like content.

23. Getting an AI

• Most people know Chat GPT from OpenAI. We use Chat GPT5.4, because it has custom GPTs inside that perform like apps.
• Many Irish companies have Microsoft 365 with Microsoft Copilot. Microsoft Copilot is owered by OpenAI.
• You can also get AI apps for your phones and desktops. We show paid versions of several AI apps in this course.
• We also demo services that power AI in web browsers. They are often available inside GPT5.4.

24. Digital Transformation

• It is a key performance indicator for hundreds of Irish businesses (CIPD).
• Embracing digital technologies is crucial for a competitive, innovative and resilient enterprise (Dept. ET&E)

25. Digital Transformation

• Developing digital-based skills – becoming digitally mature.
• Key performance indicator for hundreds of Irish businesses (CIPD)
• Embracing digital technologies crucial for a competitive, innovative and resilient enterprise (Dept. ET&E)

26. Personal Digital Transformation is…

… leveraging technology to improve your productivity, knowledge, skills, and abilities. … harnessing the power of digital tools to inform the decision-making process. …. using digital technologies to allow us, our people, and organisations develop and grow.

27. … being digitally mature

“In essence, digital transformation is about an organisation and their people’s ability to adapt to rapid changes caused by evolving digital technologies…. Digital transformation is about becoming more digitally mature as an organisation.”

(Note: digitally immature: think Blockbuster, Kodak, Ericsson, Blackberry)

Source: CIPD, 2021Digital transformation in organisations and people functions

28. Be aware of complexity

“Long-term forecasting is almost impossible for chaotic systems, and dramatic change can occur unexpectedly; as a result, flexibility and adaptiveness are essential for organisations to survive”. – David Levy, 1994

• David Levy, 1994

29. Be sceptical

• There are patterns to the data, but also large gaps between the data points.
• Data cannot always predict the strength and direction of this change.
• We don’t know what’s coming.
• Stay sceptical but curious and flexible.

30. Digital Transformation - Trust

• Deep Fakes
• AI generated articles & reports
• Propaganda
• Misinformation/Disinformation
• Hype: both positive & negative
• Control in the hands of only a few

31. AI’s biggest risk

“AI’s biggest risk isn’t ‘consciousness’—it’s the corporations that control them.”

(Meredith Whittaker, 2025)

32. The Four-P Framework

• Purpose
• Programs
• People
• Protocols

33. Purpose

34. Purposeful Transformations

• Filtering information
• Prioritising notifications
• Automating repetitive digital tasks
• Quickly producing actionable meeting minutes
• Researching & Content Creation

35. Workflow to leverage AI tools

Using AI to generate content and cross-publish

36. Use AI to keep you updated

37. Use Case: use Claude to research

39. Start with Claude settings

Adjust your privacy settings:

Profile > Settings > Privacy

40. Claude Custom Instructions

Tell Claude how you want it to respond

Profile > Settings > General

41. Create projects with Claude

Use Projects to organise your work

Home screen > Left menu > Projects > new Project

42. Upload Resources

• Use Projects to organise your work
• Home screen > Left menu > Projects > new Project
• (Upload research files you want Claude to read and give it project specific instructions)

43. Four Part Framework: People

• Purpose
• Programs
• People
• Protocols

44. Keep People Front & Centre

What is “Artificial” Intelligence?

•      Intelligence is more than cognition (Gardner MI, Kaufman 4Cs, Sternberg Success) •      AI reads and creates text, numbers, images (predefined parameters) •      Apps are efficient and fast (cognitive) •      Apps are not autonomous (yet) •      Apps need instructions from humans

45. On the question of digital transformation

On the question of digital transformation…

“The answer is less technological than you may think."_Kane et al., (2019) The Technology Fallacy: How People are the Real Key to Digital Transformation

46. Some Key Findings of Kane’s Research

A collaboration between MIT Sloan & Deloitte Digital found that…

•          Digital maturity is more about people than technology. •          A culture that embraces change, measured risk-taking, and innovation is vital for survival and growth. •          Continuous learning and adaptability are necessary. •          Organisations should encourage experimentation. -          New digital technologies disrupt established business models. •          Flexibility in organisational and individual responses to digital changes is crucial

More: The Technology Fallacy

47. People and skills needed

• Coping with change – engage them in the process
• Selection & Assessment – human decision-making
• Learning & Development – training staff
• Leadership – focus on culture

(Institute for the Future of Work, 2026)

48. Keep people front and centre

49. Four Part Framework: Protocols

Sophisticated Digital Transformation leverages Application Programming Interfaces.

• Expect data transfer between apps.
• This is happening fast.
• Future IRDG sessions offer specifics.

50. What tools and how do they integrate

How tools, programs, and applications work together.

• For personal digital transformation
• For collaboration with people
• For content creation

51. Protocols facilitate this workflow

52. Protocols connect applications

•Sophisticated Digital Transformation leverages Application Programming Interfaces (APIs).

•Expect data transfer between apps

•A contract to share data between applications

•What data are you sharing and how is it being used?

53. Voice is a protocol

Natural language forms part of this app data sharing since 2014 or even earlier

• Apple Siri
• Amazon Alexa
• Google Assistant

54. These tools are not new

• Siri launched in Oct 2011
• Google & Bing Search are AIs
• Supermarket apps are AIs
• Social networks run AIs
• Spotify, Apple Music, etc. run AIs
• Predictive text on your phone is an AI
• Verbal discussion with generative AIs

55. People use personal LLMs as a protocol

• Your company may opt to build its own local Large Language Model (LLM).
• Bernie uses an Open Source LLM with Obsidian. No internet connection required.

Or use a trusted enterprise service LLM (Microsoft or Google services

56. Consider energy usage of AI

• A ChatGPT query that produces video takes 100x more energy to execute than a Google search query
• When training Llama 2 (modelled after ChatGPT), Meta burned 539 metric tons of CO2 Larger language models use more energy during their deployment

57. Prompt engineering framework

The objective of a prompt is to provide the AI with boundaries and context. If we fail to put reins on the AI, it’s response will be less than optimal.

58. Why GCAO works

• It aligns with research indicating that clearly defined objectives boost LLM performance.
• It ensures the model has the required frame and structure to produce usable output — particularly useful given your focus on schematics, data-driven workflows, and well-structured deliverables.
• It’s straightforward and repeatable; you can embed it as a template in your prompt‐engineering workflow.

59. Goal

• Specify what you want to achieve.
• For example: “I want to reduce the time it takes to process our sales data and produce quarterly reports.”

60. Context

• Provide background information the model needs.
• Example: “I am a sales manager responsible for producing quarterly sales reports for management. We are an engineering company providing bespoke machinery to the farming sector. Data is spread across a number of applications including Google Drive, CRM, and local folders.”

61. Action

• Define the task you want the model to perform.
• Example: “Guide me step-by-step and help me produce the quarter the sales report with data visuals and summary. Review all the materials cited in the attached files. Confirm you can open and view all the materials identified in this prompt. Let me know if you are unable to access any of the source materials. Ask me questions to help improve the output.”

62. Output

• Describe how you want the answer to be structured.
• Example: “Create an interactive dashboard of all the relevant data points that have been extracted from the source materials used to produce the dashboard. List titles of the source materials as an appendix to the dashboard.”

63. Breakout 2: Ask AI about this training.

• Download the CourseSummary.PDF
• Upload the PDF to your AI and prompt using the GCAO prompt framework.
• Find out how this programme can help you in your work.
• Put your prompt into Zoom chat.

64. Perspectives and Recap

65. Bias beware

• Every AI can generate biased results.
• You have oversight and supervisory responsibilities.

66. AI and jobs

67. AI adoption increases unemployment

1. Verizon plans to lay off 13,000 employees.
2. Boston Consulting Group believes 10% to15% of U.S. jobs could be eliminated within five years.
3. Loss of those jobs will reduce purchasing power from an economy heavily reliant on consumer spending.
4. There is a hot debate over whether new jobs will be created, and if so, when.

(BCG, 2026)

68. AI and Jobs

It’s not always AI itself that forces change.

69. Trustworthy AI

According to EU AI Guidelines, trustworthy AI should be:

1. Lawful -  respecting all applicable laws and regulations
2. Ethical - respecting ethical principles and values
3. Robust - both from a technical perspective while taking into account its social environment

(European Commission, 2024)

70. Ethics and AI

Ethics guidelines for trustworthy AI

“AI systems should empower human beings, allowing them to make informed decisions and foster their fundamental rights. At the same time, proper oversight mechanisms need to be ensured, which can be achieved through human-in-the-loop, human-on-the-loop, and human-in-command approaches“.

(European Commission, 2024)

71. EU AI Act will collide with USA

• DOGE used AI to lay off people ad to reduce on-site quality controls.
• AI is replacing funded American scientific research.
• Facial recognition used by US ICE to identify immigrants.
• Streamlined national agencies now using chatbots to replace call centre staff, algorithms to replace personnel while laying off tens of thousands of federal workers, technologists, scientists, and civil servants.

72. Bots outnumber humans online

73. You will use Google Gemini

•Gemini can “read” across your Google apps to save you time. •Gemini can browse the web for 20+ minutes, analyze hundreds of sources, and generate a multi-page, cited report while you do other tasks. •Video to text, image generation, and video generation Streamlined national agencies now using chatbots to replace call centre staff, algorithms to replace personnel while laying off tens of thousands of federal workers, technologists, scientists, and civil servants.

74. You will use Claude

•Automates tedious tasks. Describe what you need. It does the work.

•No coding required. Just use plain English instructions.

•Works with your existing files. Documents, spreadsheets, folders—directly.

•Fewer errors. Consistent execution (with context window exceptions

•More time for real work. Less admin. More expertise-driven tasks.

75. Claude.ai. Claude Cowork

• Operates Locally
• Helps you manage and organise
• Take actions on your behalf
• However, has less flexibility and scope than Claude Code

76. What we did today

1. Saw four AIs create content for this session.
2. Learned about five common AI tools.
3. Saw Claude.ai write localised content.
4. Used AI to find out info about this training course.
5. Questioned the role of AI for productivity.

77. Why you are here

•You are here to learn how to use AI for stress tests with your research. •You are learning how to show your AI illiterate colleagues how to raise the bar of productivity.

78. AI tools in this course

•Chat GPT 5.5 •Claude.ai •Google Gemini •Microsoft Copilot

80. Primary Sources

• CIPD (2021) Digital transformation in organisations and people functions
• Institute for the Future of Work. (n.d.). Our work. https://www.ifow.org/our-work 
• Kane et al., (2019) The Technology Fallacy: How People are the Real Key to Digital Transformation
• Levy, D. (1994). Chaos theory and strategy: Theory, application, and managerial implications. Strategic management journal, 15(S2), 167-178.
• Mollick, E. (2023) Co-intelligence.
• Pratschke, M. (2024) Generative AI in Education

81. References

•Cybernews. (2026). Poll finds Americans truly despise AI – even more than ICE. https://cybernews.com/ai-news/poll-americans-despise-ai-ice/

•Gizmodo. (2026). People Hate AI Even More Than They Hate ICE, Poll Finds. https://gizmodo.com/people-hate-ai-even-more-than-they-hate-ice-poll-finds-2000731438

•MindStudio. (2026). What Is the AI Backlash? Why Public Sentiment Toward AI Is Worse Than ICE. www.mindstudio.ai/blog/ai-b…

•Quinnipiac University. (2026). The Age Of Artificial Intelligence: Americans' AI Use Increases While Views On It Sour. poll.qu.edu/poll-rele…

•Stanford HAI. (2026). Public Opinion | The 2026 AI Index Report. https://hai.stanford.edu/ai-index/2026-ai-index-report/public-opinion

•Boston Consulting Group (2026). AI will reshape more jobs than it replaces.

82. Links to videos

• Planning AI for Personal Productivity
• How Transformers Work - Claude helps update slide deck
• Tools that enhance productivity

83. Next session

• Frameworks for projects
• Areas of effort
• Resources
• Leveraging your archive
• Managing context

84. Thanks

85. Join us at Amplify 2026

June 18, 2026