Human-Centric AI Implementation for CGLCC Community
Ops Machine
Our Philosophy
We believe the future of AI is deeply human.
AI Amplifies Human Potential
AI Replaces tasks, not humans.
Every Leader's Vote Counts
But mostly in proportion to our level of success
Good People Must Lead
It's critical that ethical leaders embrace AI and succeed
👋🏼 Hello there!
Mitch Schwartz, Founder & CEO of Ops Machine
CGLCC-certified consultancy
AI Consultant, Process Designer, Startup Founder, Business Owner, Product Developer & Designer, Agile Coach, yada yada…
With years of expertise in process design and team dynamics…
We help solo, small & mid-sized biz's adopt AI intentionally
Measurable AI wins—balancing automation with human impact.
What You'll Learn
Finding how to integrate AI
Avoid common pitfalls
Align AI to your team, organically
Ethics and Bias 101
Section 1: Human-Centric AI Interaction
Good News Everyone!
You Already Have the Skills
AI isn't super technical.
It's much more like talking to a person.
You can:
Yammer at it
Ask Questions
Provide or ask for structure
Keys
Process Design
Make diagrams of your workflows!
Intuition
Play with it, build an intuitive feel
Be a Leader
You're guiding a book-smart intern
The Myth About AI Skills
Myth →
Overthinking
AI requires technical expertise or complexity
Underthinking
AI is hype but doesn't work for me
Reality
Most of us already have the base skills
Good results need come from building AI muscles.
AI Evolution Model
Where to Start:
Level 4: Automation
Systems integration
Level 3: Custom GPTs & Master Prompts
Specialized Contexts
Level 2: Context-Informed
More Consistent Results
Level 1: Raw chats
Get Started
Start at Levels 1-2 for most value
The Power of Level 1: Conversational AI
Even with "Raw Chats", you can get a lot.
Treat AI as a Colleague
Engage in a dialogue as if with
a human team member.
Ask for Frameworks
Before asking for answers or even outlines,
ask for best practices and frameworks
Seek Feedback & Opinions
Ask clarifying questions
Ask "What do you think?"
Level 1: Being a good AI conversationalist
Unfocused
"Write a blog post About using AI"
Context
"I want to write a blog post for my business which helps people learn AI skills and adopt AI so that they can become more productive. What are some of the best practices right now in terms of interacting with AI? Please search for articles only written after Jan. 2025"…
"Great, now what are the top frameworks and techniques that you've found, and what are the most fundamental skills? Also, how can we really make this valuable and not just generic, I hate generic LinkedIn trash posts"
Level 2: The RICO Model
RICO Model: A framework for effective AI interaction
Role
Define the AI's expertise and perspective
Instructions
Clear directions on what you need
Context
Provide relevant background information
Output
Specify the desired format and detail level
Exercise 1
Identify your AI-ready workflows (or steps):
What recurring work drains your energy?
What parts of this process have repeatable steps?
What specific steps are most time consuming or draining?
How Would someone new to your company learn how to complete these Steps?
Section 1 Insights
Energy-draining workflows = AI opportunities (if processes are clean)
Start with conversation (Level 1-2), not automation
Human-centric approach beats commands
Example: 3-bot "automation" built by conversation
I created 3 bots in ChatGPT to run user interviews and update personas.
1: Question Maker
Reads personas and creates interview questions.
2: Interviewer
Leads the user interview.
Bot 3: Updater
Reads interview transcripts, identifies personas, and updates.
The Approach
Explain Bot's Role
Ask if it has questions about how to do the role
Answer the questions, ask if it has new questions
When no more questions, ask for updated prompt
Excerpt from Persona Update
2SLGBTQIA+ Founder Goals & Aspirations:
Build workplaces and systems that reflect personal values and decolonized structures:
non-hierarchical, flexible, and inclusive environments.
Focus on equity and impact:
Grow businesses that uplift marginalized groups or serve broader social justice missions.
Balance personal well-being and advocacy:
Demonstrate that success is possible without sacrificing core identities and ethics.
Section 2: Human-Centric AI Implementation
Workflow Selection Process
"Before you automate, optimize" (Don't delegate messy process)
Step 1: Energy Drain Audit
- "What task makes you think 'ugh, not this again'?"
- "What do you spend time on that doesn't require your specific expertise?"
- "If you could wave a magic wand and never do one thing again, what would it be?"
Step 2: RICO
- Role: Is there a clear model for the AI's function?
- Instructions: Can the steps be clearly defined?
- Context: Is all necessary information accessible?
- Output: Is the desired outcome measurable and specific?
Step 3: AI Suitability
- Repetitive: How often does this exact type of work happen?
- Rule-based: Are there clear decision criteria?
- Data-rich: Is the required information readily available?
- High-volume: Does this task consume significant time or resources?
Having Success with new Team Processes
Success requires deep connection with your team:
Curiosity
About how people actually work
Iteration
Loops throughout implementation
Inclusion
Of diverse voices
Clear Guidelines
So people can explore safely
Teams + Evolution Model
Levels 1-2: Foundation & Experimentation
- Teams focus on training, policy setting, and knowledge sharing.
- They experiment with AI to build comfort and a deeper understanding of its capabilities.
Teams + Evolution Model
Levels 3-4: Automation & Integration
- Focus on shared context
- 'Atomic' Automations based on proven successes.
- Human-in-the-loop at every level, for oversight and validation.
Master Prompts
AI's Key Context
Unified Context
Store key company knowledge and processes
Eliminates Retraining
AI has core details without repetitive input
High-Quality Output
Relevant, accurate, and more consistent results
Quick Analysis
Faster insights and better decision-making
Master Prompt in Action
Sitting on a downtown bench
15 mins early for an event, I realized I hadn't reviewed the attendee list.
On my phone
I shared the list into our a new chat, in a project with our Master Prompt.
In 3 Minutes, I ID'd and ranked my top prospects
with brief explanations for each.
Policy Basics
One critical pitfall is the absence of a clear AI policy.
Without proper guidance, teams can stumble and create risks.
Data Usage Guidelines
Define what type of data is and isn't permissible to share with AI models.
Approved Tools
Clearly list which AI tools are sanctioned for company use.
Tool Request Process
Establish a clear process for requesting permission to use new AI tools.
Prohibited Actions
Specify actions, behaviors, or data inputs that are strictly off-limits.
Policy Examples
Concrete policies guide responsible AI adoption within your team:
Controlled Data Usage
Use shared, data-protected AI accounts not Individual accounts.
Mandatory Human Review
All AI-generated work must be human-reviewed.
Bias & Ethics Checks
Implement a process for reviewing all AI-generated content for potential biases.
AI Exploration Checklist
Plan your team approach:
Are you involving people who do the work?
Are you providing guidelines and boundaries?
Are you providing tools and training?
Are you working iteratively and actively fostering feedback?
Are you keeping humans in the loop?
Section 2 Insights
Keep Humans in the loop! Focus on Iteration.
Teams need to evolve together through AI levels.
Inclusion, feedback, boundaries, and clarity prevent many pitfalls.
Section 3: Human-Centric Ethics
Human-Centric Ethics
- EU AIA provides roadmap for everyone
- Bias is part of the system
- Human oversight critical for community trust
Environmental Impacts
There's no simple answer, but we can strive for more sustainable practices.
Lean Model Selection
Prioritize using the lightest useful models for your task
Efficient AI Use
Optimizing interactions: build your skills to do more, faster
Artificial Intelligence Act
The EU's (AIA) introduces a risk-based approach to AI regulation.
Minimal Risk
e.g., video games, basic filters
no specific rules.
Limited Risk
e.g., chatbots, deepfakes
disclosure required.
High Risk
e.g., recruitment, health diagnostics
full compliance package & human-in-the-loop controls.
Unacceptable Risk
e.g., subliminal manipulation, social scoring
outright banned.
Ethics & Bias
The AIA addresses ethical considerations and bias:
Data Governance
Relevant, representative, and error-free datasets
for training AI models.
Risk Management & Oversight
Embedding bias checks, continuous monitoring, and corrective actions throughout the system's lifecycle WITH human oversight.
4-Point Ethics Framework
Risk Check
High-stakes decisions needing compliance?
Bias Check
Tested with diverse scenarios (including LGBTQ+)?
Transparency Check
Understand it + human oversight?
Values Check
Diverse voices involved + aligns with values?
Justice AI
Decolonial AI
Trained on decolonial datasets.
Bias Audits
Can identify bias and discriminatory practices.
Safe Learning
Provides a safe, space for educational interactions
Audits & Consulting
Justice AI Also provides consulting services
Exercise 3
Design your ethical pilot:
How will you run checks for: Risk, Bias, Transparency, and Values?
How can you test different models for your use case?
Section 3 Insights
Workflows can be more ethical with human-centric approach
Ethics woven throughout, not bolted on
Curiosity + feedback + inclusion prevent bias
Key Takeaway 1
Human-centric AI interaction is conversation, not commands
- Use process design + conversational AI
- Start Levels 1-2, not complex automation
- Internalize a feeling for working with AI
Key Takeaway 2
Human-centric implementation prevents pitfalls
- Deep connection (curiosity + feedback + inclusion) with team
- Provide clear guidance
Key Takeaway 3
Human-centric ethics is woven throughout
- 4-point framework at each level
- Diverse voices + human oversight = values alignment
AI Labs & Mastermind
Join our FREE weekly sessions to deepen your practical AI knowledge
Every Wednesday, 1:00 PM Eastern.
Ask questions, share ideas, learn in a grounded, way.
Free and open to anyone interested in AI.
Exclusive 2SLGBTQIA+ Community Offer
4-Hour "SpAIral Up" AI Jumpstart
From "Where Do We Start?" to Measurable AI Wins
What you get:
4-hours of consultation time solving how AI advance YOUR business.
Pricing:
Pay What It's Worth (After You See the Results)
More info: opsmachine.co/cglcc-offer
Adopting AI: Practical, Ethical, and Scalable Paths Forward