AI is About People

This week marked a milestone for me.

The successful launch of All Things AI (formerly AllThingsOpen.AI), a conference I co-founded with Todd Lewis, the creator of All Things Open, the Southeast's premier tech conference for open source enthusiasts.

Our inaugural event in Durham, NC exceeded all expectations:

👔  1500 Unique Conference Registrants
👩‍💼  750 Business Users Trained on Generative AI 𝗟𝗶𝘃𝗲
👍  100% of training respondents would recommend to a friend
💫  Over 70+ 4.7/5 Star Reviews for our training
🔉   50 World Class Speakers

The remarkable turnout confirmed something I've observed: AI uniquely transcends traditional boundaries, drawing interest from people across all industries, backgrounds, and walks of life.

Despite the technical focus, what stood out most was the fundamentally human dimension of AI adoption.

Never before has a technological revolution sparked such universal curiosity and need for guidance. The success of this event reinforced our mission: providing accessible resources to help everyone understand and harness AI's potential, regardless of their background.

So here’s a recap of lessons I learned, applications I saw, and some advanced prompting to help me with the planning of next year’s event.

☕️ AI Tangle - Apple Intelligence's Lawsuit Troubles, World's First AI-Generated Newspaper & Meta AI's Release in Europe

 🔮 AI Lesson - Creating Killer Presentations with ChatGPT 

🎯 The AI Marketing Advantage - Stop Your AI Agents From Resetting—Give Them Memory

💡 AI CIO - Promise and Perils of AI Development

 📚 AIOS - This is an evolving project. I started with a 14-day free Al email course to get smart on Al. But the next evolution will be a ChatGPT Super-user Course and a course on How to Build Al Agents.

AI is About People

With artificial intelligence, we need to focus on the people as much as we do the technology

When I got into AI one of the first things I did was format a meetup group to discuss what was going on with others. (We do both live and online events for those not near Raleigh-Durham, NC)

This week, I took that meetup group to the big time with the conference, All Things Open AI (now just All Things AI or even shorter ATAI.)

From the opening remarks to the final sessions, openness was the defining theme of ATAI. “This was the only [AI conference] focused entirely on openness,” noted one attendee​

IBM’s Sriram Raghavan, Vice President of AI Research, drove this point home in his Tuesday keynote: “We truly have been committed to an open approach to AI… at the platform level and increasingly when it comes to models, agents, and all of the new things happening with AI,” he said. “Having a conference in which we can talk about what we are doing – and share [it] with the community – is fabulous.”​

Raghavan then delivered one of the event’s biggest announcements: IBM is contributing three popular AI developer tools to open source under the Linux Foundation.

This commitment to openness was echoed widely. “Open source models are proving competitive with proprietary solutions, accelerating industry-wide innovation,” wrote DevOps veteran John Willis (John also publishes our AI CIO publication). He’s been advocating as I have, using IBM’s Granite models to train bespoke models on enterprise data.

Speakers from tech giants like Google Cloud, Netflix, Walmart, Fidelity, and Red Hat likewise shared how they leverage open-source AI in practice.

Even first-time attendees could sense the ethos: “It was exciting — just feeling that buzz of innovation in the air,” one participant remarked after two days of open AI discussions​

Security, Trust, and Responsible AI

Open innovation didn’t mean ignoring risks – AI security and ethics were front and center in many sessions. One takeaway was that embracing open source goes hand-in-hand with responsible AI. Willis warned of “Shadow AI,” the scenario where developers use AI tools under the radar, potentially exposing organizations to compliance and security risks.

He argued that truly open AI requires full transparency of code, data and model weights – not “selective openness” – so that users can trust and verify the technology.

Security and governance matter, enterprises must institute robust safeguards (for example, defending against prompt injection “jailbreaks”) and align their leadership and compliance teams to manage AI responsibly.

Concrete security solutions also made an appearance on the expo floor. StackLok showcased its new CodeGate platform, which helps developers stay secure when using AI coding assistants like GitHub Copilot. The tool scans AI-generated code for vulnerabilities, addressing a key concern as AI enters daily development workflows.

Meanwhile, keynote speaker Dr. Ruth Akintunde of SAS highlighted strategy over fear. She advocated a “hybrid approach to AI,” suggesting that businesses blend multiple AI techniques (and keep humans in the loop) to get the best outcomes.

Leveraging AI’s power while maintaining oversight – resonated with many in highly regulated industries. Taken together, these talks on AI trust, security, and governance added a sober note to the AI enthusiasm, ensuring that “open” also means accountability and safety.

Workshops, Demos and Tech Highlights

The conference wasn’t just talk – it was very much about building with AI. Day 1 was a hands-on training day, where attendees packed two in-depth workshops: an “AI SuperUser Bootcamp” and a “GenAI RAG (Retrieval-Augmented Generation) Workshop”​

In these sessions, I led my AIOS SuperUser Bootcamp, where participants learned advanced AI techniques, prompt engineering, AI integration techniques, and AI integration techniques, illustrated with real-life examples.

“Serious learning” was the goal, and the energy continued into Day 2’s demos and lightning talks​

Several live demos earned buzz for showing AI in action. In one talk, Six Feet Up CTO Calvin Hendryx-Parker rolled out an end-to-end demo of deploying enterprise AI using RAG (retrieval-augmented generation) to transform how internal teams find information.

Many were impressed to see a real-world example of connecting a large language model to private company data – a technique that made abstract AI concepts feel tangible. GSK and LabCorp both gave talks on how they were using AI to generate better outcomes in medical testing and drug safety.

The theme of AI agents also popped up throughout the agenda. IBM’s BeeAI, one of the newly open-sourced tools, is itself a framework for discovering and orchestrating AI agents, hinting at a future of interoperable “worker agents.”

Duke University’s AI Master’s program had a visible presence as well (even hosting a booth), reflecting how academia and industry are joining forces to push AI forward. ​From database companies (Couchbase, Elastic) to dev tool providers (GitLab, Docker) and cloud platforms, a diverse array of tech players showed off integrations and AI-powered features.

Whether for talks on semantic search, AI in fintech, or the “Future is Non-Deterministic” (a forward-looking session on AI’s unpredictable evolution), attendees were eager to see the latest AI innovations up close.

Community Buzz and Noteworthy Moments

The community reaction to ATO AI 2025 has been overwhelmingly positive. As the event unfolded, many attendees took to LinkedIn and X (Twitter) to share highlights and kudos.

Others described the content as both inspiring and grounded. The “buzz of innovation” was a recurring theme in these posts, as developers and data scientists traded their favorite takeaways – from cool demos to practical AI governance tips.

Notably, the “openness” ethos extended beyond technology to networking and inclusivity. Attendees celebrated how All Things Open AI brought together people from diverse backgrounds – developers, business leaders, academics, and open source contributors – all eager to learn from each other. ​

Here are some of my favorite comments from attendees:

In terms of surprises, IBM’s open-source trio announcement certainly tops the list – few expected a major tech company to drop news at a first-year conference, which instantly put ATO AI on the map. The huge turnout was another pleasant surprise. By official counts, around 750 people attended the workshops on Monday, and about 1,000+ filled the halls on Tuesday

As the curtain closed on All Things Open AI 2025, the sentiment was clear: this community is just getting started. I am currently getting the videos translated to share online since I think the content was so good I’d hate to keep it closed. Stay tuned, I’ll be sharing it and some more online events around All Things AI soon.

There were quite a few projects that I saw presentations on at All Things Open AI. There was a theme, lots of open source. Here are my top picks.

Obot - Obot is an open source AI assistant platform offered as a SaaS and self-hosted deployments. The Obot SaaS platform is built on top of the same open source technology that you can deploy yourself, offering flexibility in how you access and control your AI assistants. Try it, I can guarantee you wont’ be disappointed.

Bee AI - BeeAI is an open platform to help you discover, run, and compose AI agents from any framework and language. Whether building your agents or looking for powerful existing solutions, BeeAI makes it easy to find, connect, and orchestrate AI agents seamlessly.

Data Prep Kit - Data Prep Kit accelerates unstructured data preparation for LLM app developers. Developers can use Data Prep Kit to cleanse, transform, and enrich use case-specific unstructured data to pre-train LLMs, fine-tune LLMs, instruct-tune LLMs, or build retrieval augmented generation (RAG) applications for LLMs. Data Prep Kit can readily scale from a commodity laptop to data center scale.

Docling - Docling simplifies document processing, parsing diverse formats — including advanced PDF understanding — and providing seamless integrations with the gen AI ecosystem.

Granite LLM Models - Granite is a family of AI models purpose-built for business, engineered from the ground up to ensure trust and scalability in AI-driven applications.

Ollama - Get up and running with large language models running locally on your own machine.

Complex Planning with ChatGPT

I have a friend who early on in the genesis of AI asked ChatGPT, “How do I become the leading thought leader in AI?”

He then chatted back and forth until he had a plan.

I’d say the plan was a success. He’s charging hundreds of dollars for his time. He’s a trusted advisor to many large companies implementing AI. And arguably is one of the most interesting folks in AI with his latest Agentics Foundation.

His name is rUv, check him out on LinkedIn if you want to get insights on becoming a power user.

Now back to the prompt.

My goal is to have the leading conference for AI practitioners and users focused on technologies, processes, and people. So of course I am going to use AI to help inform the plan.

I had a few options to do this but I focused on one in particular.

1. I could just create a chat, upload all the data I have including all the links to all the feedback forms that are publicly available, and have ChatGPT search for the results.

2. I could export all the data and create a Project in ChatGPT and do a fine-grained analysis of the attendees in on the chat. Ticket sales in another. Feedback in another and dive into each aspect then send all the summaries to a single chat.

3. I could even create an All Things OpenAI Custom GPT and add all the reports and information to the knowledge base and then tweak the custom instructions to allow me to chat with the entire amount of data.

I chose option 3 so I can keep adding data and asking questions. For example:

• Were our ticket prices too low or too high?

• How do our sponsorships compare with other events?

• What was the overall sentiment from attendees?

Now this isn’t a cut-and-paste prompt it’s meant to show you how to analyze data and use ChatGPT to do some advanced reasoning on a complex data set.

In this example, I am optimizing for success for next year.

I uploaded all the relevant documents from this year’s event, including ticket sales data and attendee reviews. I provided links to the prospectus for sponsors and media kit.

Then I started chatting with the data to get insights. I also could share that GPT with my team. You may not be planning a conference but it’s actually a very effective way to approach any large, complex planning task.

Also, this is my specific prompt you can take this prompt and cut and paste it into ChatGPT and improve it for your specific big project.

# Objective

Develop a plan that analyzes conference data and informs the planning of future AI-focused events

# Role 
Act as a conference organizer who excels at providing world-class technical experiences while being efficient at using their budget. 

# Instructions

**1\. Data Collection and Integration**

-   **Attendee Demographics:** Collect data on attendee roles, industries, and experience levels to tailor content effectively.​

-   **Session Feedback:** Gather ratings and comments for each session to assess content relevance and speaker performance.​

-   **Engagement Metrics:** Analyze participation in workshops, Q&A sessions, and networking events to identify popular formats.​

-   **Social Media Activity:** Monitor mentions, shares, and comments to gauge real-time attendee sentiments.​

**2\. Data Analysis and Insights Generation**

-   **Trend Identification:** Detect emerging topics and technologies that resonate with attendees.​

-   **Speaker Evaluation:** Assess speaker effectiveness based on feedback and engagement levels.​

-   **Format Preferences:** Determine preferred session formats (e.g., panels, workshops) to optimize future agendas.

-   **Networking Efficacy:** Evaluate the success of networking opportunities to enhance attendee connections.​

**3\. Recommendations for Future Conferences**

-   **Content Development:** Prioritize topics with high attendee interest and satisfaction.​

-   **Speaker Selection:** Invite speakers who received positive feedback and align with attendee interests.​

-   **Session Structuring:** Adopt preferred formats and allocate time slots based on engagement data.​

-   **Networking Enhancements:** Implement strategies to improve networking based on past participation metrics.​

**Considerations for Creating an Industry-Leading AI End-User Conference**

-   **Clear Objectives:** Define specific goals to guide planning and execution.​

-   **Target Audience Understanding:** Develop detailed personas to tailor content and experiences.​

-   **Comprehensive Budgeting:** Align financial planning with event objectives to ensure resource allocation.​

-   **Engaging Agenda:** Balance educational sessions, networking, and entertainment to maintain attendee interest.​

-   **Technology Integration:** Utilize event management software and mobile apps to enhance the attendee experience.​

-   **Sustainability Practices:** Incorporate environmentally responsible measures to appeal to eco-conscious attendees.​

-   **Post-Event Evaluation:** Collect and analyze feedback to continuously improve future events.​

# Success Metrics 
- Attendance 
- Sponsor satisfaction 
- Speaker satisfaction 
- Attendee satisfaction 
- Profits
- Brand recognition 

I appreciate your support.