How Do You Teach AI Your Business?

AI already knows a lot.


That is not really the problem.


The challenge is teaching it what matters to your business.


It does not automatically know your customers. It does not know your tone. It does not know your standards. It does not know your process. It does not know which details are important and which ones will send your team into a group text spiral.


That is where the skill comes in.


Using AI well is not just about asking better questions. It is about teaching AI what you want it to know so it can give you answers that are actually useful.


That is the focus of the class I am teaching through the BBB AI Hub: Teaching AI your Business: Building Context That Actually Works.


We are going to talk about how to use ChatGPT settings, Projects, and Custom GPTs to give AI better business context.


Not in a complicated way.

In a practical way.


Because most business owners do not need more AI hype. They need to understand how to make AI useful in the work they are already doing.

Why does AI give generic answers?

Most people have had this experience by now.

You ask ChatGPT to write an email, summarize a meeting, create a post, draft a proposal section, or help organize an idea.


And the answer is fine.

Not terrible.

Not completely wrong.


Just very clearly written by something that has never met your customer, your team, your business, or your patience level.


That is what happens when AI does not have enough context.


If you ask for a client email without explaining the client, the situation, the tone, the goal, and what you do not want to say, AI has to guess.


Sometimes the guess is helpful.

Sometimes it sounds like it was written by a very confident intern on their first day.


The fix is not always a better prompt.

Sometimes the fix is better context.

What does it mean to teach AI your business?

Teaching AI your business means giving it the information it needs to understand how you work.


That might include:

What your business does.

Who your customers are.



How you talk to people.

What services you provide.

What your values are.

What your process looks like.

What kind of answer is actually useful.

What should be reviewed by a human.

What should never be promised.


This is where AI starts becoming more practical.

Not because it magically knows your business.


Because you gave it enough information to support your business.

That is a very different thing.


Where does that information go?

This is one of the most helpful things to understand.

Not all business context belongs in the same place.

Some information belongs in your ChatGPT settings.

Some information belongs in a Project.

Some information belongs in a Custom GPT.


That sounds technical, but it is actually pretty simple.


Settings are your baseline.

This is where you tell ChatGPT the basics about you and how you generally want it to respond.


For example, you might tell it what kind of business you run, who your audience is, what tone you prefer, and what kind of help you usually need.


Projects are your workspace.

This is where you can keep related work together.


A marketing campaign.

A proposal.

A client effort.

A training plan.

An internal process.

A set of documents you want to work from.


Projects are helpful because they keep the conversation, files, and instructions connected to the work.


That means you do not have to start from scratch every time.


Custom GPTs are repeatable tools.

A Custom GPT can be built for a specific purpose.

A proposal assistant.

A client communication assistant.

A content assistant.An onboarding assistant.

A project documentation assistant.

An internal FAQ assistant.


The value is that the instructions, knowledge, and purpose are already built in.


So instead of teaching AI the same thing over and over, you can create a tool that is designed to help with that kind of work.k

Why are Custom GPTs useful for teams?

This is the part I think more businesses need to understand.

Custom GPTs can be shared.

That means they are not just personal productivity tools.

They can become shared internal resources.


That matters because business knowledge should not always live with one person.

If one person has the “good prompt” saved in a document somewhere, that helps one person.


But if the whole team needs to work from the same proposal structure, the same client communication standards, the same internal process, or the same project information, a shared Custom GPT can help create consistency.

For example, a team could build a Custom GPT to help with:

Proposal development.

Internal project planning.

Client communication drafts.

Meeting summaries.

Onboarding.

Marketing review.

Project documentation.

Internal questions and answers.

That does not mean AI replaces review, strategy, or human judgment.


It means the team has a better starting point.

Everyone is not reinventing the wheel.

Everyone is not using a different version of the instructions.

Everyone is not trying to remember where the last good example was saved.


A shared Custom GPT can help people work from the same context.

And for busy teams, that is a big deal.

A simple way to think about it

Settings help AI understand you.

Projects help AI understand the work.

Custom GPTs help AI support a repeatable task that other people may need to use too.


That is it.

That is the beginning of teaching AI your business.

You do not need to learn everything at once.

You do not need to become an AI engineer.

You do not need to build the most complicated system in the world.

You just need to start giving AI better information in a more organized way.


Why does this matter for business owners?

Most businesses already have the information AI needs.

It is just scattered.

It is in emails.

It is in meeting notes.

It is in proposals.

It is in project folders.

It is in old documents.

It is in someone’s head.

It is in the way your best employee explains something without even thinking about it.


That information is valuable.


AI can help you use it better, but only if you learn how to share it clearly.


That is what makes AI skills practical.


The goal is not to use AI because everyone is talking about AI.


The goal is to use AI to make real work easier.

Better first drafts.

Clearer communication.

More consistent processes.

Faster planning.

Stronger internal knowledge sharing.

Less starting from scratch.

That is the opportunity.

What will we cover in the class?

In Teaching AI your Business: Building Context That Actually Works, we will talk about how to make ChatGPT more useful by teaching it the right information.

We will cover:

How to use settings.

How to think about Projects.

How to use Custom GPTs.

How Custom GPTs can support team collaboration.

How to share business context.

How to get better outputs.

How to keep human review in the process.


The class is introductory and practical.


No jargon marathon.

No pretending AI is magic.

No pressure to become a technology expert overnight.


Just a clearer way to understand how to make AI useful inside your business.


What is the best way to make ChatGPT useful for business?

The best way to make ChatGPT useful for business is to teach it the context that matters.

Use settings to explain your baseline preferences.

Use Projects to organize ongoing work.

Use Custom GPTs to create repeatable tools that can support specific tasks and, when appropriate, be shared with your team.


AI may know a lot.


But your job is to teach it what matters here.

Your business.

Your customers.

Your voice.

Your standards.

Your goals.

Your work.


That is how AI becomes useful.


That is what we will be learning together through the BBB AI Hub.


Ready to make AI more useful inside your actual business? Register for my June 29th BBB AI Hub class, Teaching AI your Business: Building Context That Actually Works, through the BBB AI Hub Programs & Classes page.

Share this article

Recent Posts

By Jonathan Liebert July 15, 2026
Most industrial organizations are not starting from zero. They are starting from systems that already work—just not in ways that scale easily. Legacy workflows often represent years of optimization under real constraints. They are imperfect, but they are functional. This is why modernization is rarely a technical problem alone—it is a continuity problem.  The goal is not to replace what works. The goal is to reduce friction while preserving operational stability.
By Jonathan Liebert July 15, 2026
Most industrial organizations are not starting from zero. They are starting from systems that already work—just not in ways that scale easily. Legacy workflows often represent years of optimization under real constraints. They are imperfect, but they are functional. This is why modernization is rarely a technical problem alone—it is a continuity problem. The goal is not to replace what works. The goal is to reduce friction while preserving operational stability.
By Jonathan Liebert June 12, 2026
Anthropic’s new Claude Fable 5 may be one of the most important AI model releases of the year, not just because of what it can do, but because of what it signals about where artificial intelligence is heading. Fable 5 is part of Anthropic’s new Mythos-class family of models, which sits above its previous Claude Opus models in capability. In plain English, this is not just a better AI model. This is a much more powerful system designed for complex work: software engineering, advanced research, long document analysis, scientific reasoning, visual understanding, and agentic workflows that can run across multiple steps. For business owners, that should get your attention. The promise of Fable 5 is significant. It can work on longer, harder, more complex assignments with greater endurance than earlier models. That means it may be useful for drafting strategic plans, reviewing large policy documents, analyzing contracts, supporting technical projects, building internal tools, developing training materials, summarizing research, or helping a team reason through complicated business decisions. This is the kind of AI that starts to feel less like an assistant and more like a high-level analyst. For small businesses, that is exciting. Many owners do not have a research department, a software team, or a strategic planning office. A model like Fable 5 could help smaller organizations access capabilities that used to be available only to large companies with deep benches of talent. That is the optimistic view. But there is another side to this release that small business owners should pay attention to: cost and access . Fable 5 is currently available in Claude for a limited time, with access expected to end on June 22, giving users a short window to experience one of Anthropic’s most advanced public models before access changes. But long term, this type of intelligence will not be cheap. Anthropic’s listed pricing for Fable 5 is reportedly $10 per million input tokens and $50 per million output tokens . For large enterprises, that may be manageable. For small businesses, nonprofits, and solo entrepreneurs, that pricing could quickly become a serious barrier. In everyday terms, this means Fable 5 may be affordable for an occasional high-value project, but expensive for constant daily use. Asking it to review one large report may cost only a few dollars. But using it all day across employees, documents, drafts, agents, and revisions could become a meaningful monthly expense very quickly. The concern is that Fable 5 is roughly double the cost of Claude Opus 4.8. That tells us something important: as models become more powerful, the best AI may not simply become cheaper. The market may split between everyday models for routine tasks and premium frontier models reserved for complex, high-value work. That may be the bigger story. For the past few years, small businesses have benefited from an incredible moment in technology: advanced AI was suddenly available through relatively affordable monthly subscriptions. A local business owner could access tools that felt almost magical for the price of a software subscription. That created a sense that increasingly powerful AI would always become cheaper and more accessible. Fable 5 raises a different possibility. What if the most powerful models become premium tools that only larger organizations can afford to use regularly? What if small businesses get access to “good enough” AI, while major corporations use the most advanced models for strategy, automation, coding, research, product development, and decision-making? That would be a major shift. Instead of AI closing the gap between small and large businesses, it could begin widening it again. Larger companies would have access to better reasoning, better automation, better agents, and better strategic support. Smaller firms might be forced to ration use, rely on cheaper models, or avoid advanced workflows because the cost is too unpredictable. This does not mean small businesses should ignore Fable 5. Quite the opposite. If you have access to it, test it. Use it on high-value work where quality really matters. Ask it to analyze a strategic plan, compare vendors, review a complicated proposal, summarize a large report, or help think through a major decision. Do not waste a model like this on routine emails or simple brainstorming. Use your most powerful AI where the return justifies the cost. There are also safety considerations. Anthropic has emphasized that Fable 5 includes safeguards around high-risk topics such as cybersecurity, biology, and chemistry. That matters because as AI becomes more capable, the risk of misuse also grows. Businesses should appreciate that the industry is taking safety seriously. But safety systems can also affect usefulness, especially when models become more cautious or restricted in certain areas. So where does this leave small businesses? The best approach is strategic adoption. Start by understanding which AI tasks are truly valuable for your organization. Use lower-cost models for routine work. Reserve frontier models like Fable 5 for complex, high-value projects. Train your team to understand when powerful AI is needed and when it is not. Track cost. Protect sensitive data. Keep humans involved in final decisions. Claude Fable 5 gives us a glimpse of the next era of AI: more powerful, more autonomous, more capable — and likely more expensive. That does not mean small businesses are out of the race. But it does mean they need to become smarter buyers and better managers of AI. The future of AI may not simply be about who has access to the best model. It may be about who knows when to use it, how to use it, and whether the value justifies the cost. That is the real AI Advantage. About the Author Jonathan Liebert is CEO/Executive Director of the Better Business Bureau of Southern Colorado, an AI thought leader and an adjunct professor at the University of Colorado Colorado Springs. He is the author of Thought Partner , which explores how leaders can collaborate with AI to improve decision-making and strategy. Jonathan also leads AI education and training programs through BBB of Southern Colorado to help businesses build practical AI skills for the modern marketplace.
Why OpenAI’s New Warning Matters for Small Business
By Jonathan Liebert June 10, 2026
Why OpenAI’s New Warning Matters for Small Business
June 9, 2026
Teaching for Tomorrow: Closing the expectations gap What happens when teachers are asked to do the impossible? New research from the Walton Family Foundation and Gallup examines how teachers experience their job demands. The Teaching for Tomorrow study reveals just how much unrealistic expectations and lack of role clarity impact key outcomes like job satisfaction and intention to remain in the classroom. While the research finds that many teachers face unrealistic or unclear expectations, it also confirms that with improved job expectations, teachers thrive.
Human Judgement Still Matters
June 8, 2026
AI is a powerful tool - but when it comes to values, people and consequences, human judgement still matters.
June 5, 2026
One of the best places to begin your AI journey — or refresh what you already know — is the Everyday AI podcast’s Start Here Series . This resource does a great job breaking down the basics of artificial intelligence in a way that is clear, practical, and easy to understand. Whether you are just getting started, trying to fill in the gaps, or looking for a simple way to stay current with how AI is changing work and business, this series is a helpful guide. It takes complex ideas and makes them accessible, giving listeners a strong foundation for understanding how AI tools can be used thoughtfully and effectively.  For small business owners, nonprofit leaders, and professionals who want to better understand AI without getting overwhelmed, this is a great place to start!
AI Tokens and The Economics of AI
June 5, 2026
AI Tokens and The Economics of AI how AI pricing actually works and how small businesses can get the most value from it
By Jonathan Liebert May 29, 2026
Anthropic’s newest Claude Opus 4.8 model is another reminder that the AI race is not slowing down. In 2025, ChatGPT and Gemini often dominated headlines. However, in 2026, Claude has quietly become one of the strongest AI platforms for reasoning, writing, coding, analysis, and complex knowledge work. This year, the headlines have mostly been about Anthropic. For small businesses, the most important takeaway is this: Claude is becoming less like a chatbot and more like a serious business collaborator. One of the biggest upgrades is stronger reasoning control. Claude Opus 4.8 gives users more ability to adjust how much effort the model puts into a task. That matters because not every job requires the same level of AI horsepower. A quick email draft does not need the same deep reasoning as a contract review, strategic plan, financial analysis, or complex customer research project. Better controls allow businesses to match the model to the task. The second major feature is Claude’s growing strength with agentic workflows. In Claude Code, the model can now plan larger projects, break them into smaller pieces, and run multiple sub-tasks in parallel. For technical teams, that could help with software updates, code reviews, migrations, testing, and documentation. For non-technical business leaders, it signals where all AI tools are heading: toward systems that can manage complex work, not just answer questions. The third feature is deeper collaboration. Claude has become especially useful for long documents, policy reviews, grant writing, board reports, training materials, and strategic planning. It is strong at taking messy information and turning it into something structured, thoughtful, and usable. But there is a caution. More powerful AI can also mean higher usage costs, faster rate-limit issues, and greater risk if employees use it without clear guidance. Businesses should not simply turn these tools loose on sensitive data or expensive workflows without a plan. Clade is one of the most expensive models out there, so you need to make sure you are using it correctly. The smartest approach is simple: start with high-value, low-risk tasks. Use Claude to draft, summarize, analyze, and organize. Keep humans involved in final decisions. Track cost, quality, and time saved. Claude Opus 4.8 shows where AI is heading: more capable, more collaborative, and more embedded in real work. For small businesses, that is a major opportunity — if used wisely. About the Author Jonathan Liebert is CEO/Executive Director of the Better Business Bureau of Southern Colorado, an AI thought leader and an adjunct professor at the University of Colorado Colorado Springs. He is the author of Thought Partner, which explores how leaders can collaborate with AI to improve decision-making and strategy. Jonathan also leads AI education and training programs through BBB of Southern Colorado to help businesses build practical AI skills for the modern marketplace.