Updates on AI for Community
We’ve been busy! May has been a full month of trainings, and we’re continuing to add more.
- Coming up on May 28th (that’s this month, so mark your calendar!): our free online meetup at 10am Pacific, where I’ll be showing you how to use AI to create presentations — including deep research, brainstorming, outlining, and building a PowerPoint deck using Claude. Register at aiforcommunity.org/online-meetup-registration.
- June 11th: the next free online meetup, also at 10am Pacific. I have some ideas for topics here, but would love your input! Shoot me an email if there’s something you’d like me to explore in depth.
- June 16th & 17th: our AI In Depth online training, 9:30am to 11:30am Pacific each day. This is the more advanced course — if you’ve been through AI Essentials, this is your next step. You’ll need at least one paid AI plan (ChatGPT, Gemini, or Claude) to participate. Register at aiforcommunity.org/online-training-registration.
- July 8th & 9th: our AI Essentials online training, 9:30am to 11:30am Pacific each day. This is a live, hands-on introductory four-hour class built specifically for nonprofit workers – no technical background required. Register at aiforcommunity.org/online-training-registration.
- Returning students and referrals get 25% off: just reach out to mefor the coupon.
- We’ve added new content to our resources page: check out Customizing Claude in AI Examples and Anthropic’s System Promptin Fun Stuff.
- If you know anyone at a community foundation or large nonprofit, please introduce them to us. The best way is to have them schedule a Zoom call with me.
Hot Off the Press
- AI backlash is building, and it may get ugly. The Atlantic published a piece this month with that blunt headline, and it’s worth reading. The argument is that public frustration with AI is growing — the disruption is visible, the benefits often aren’t — and that the backlash could get politically charged if layoffs and displacement start hitting harder. For nonprofits, this is a heads-up that how you talk about AI in your community matters. Being transparent and honest about what you’re using it for, and why, is good practice regardless.
- ChatGPT Images 2.0 launched. This is an upgraded image generation with better text rendering, higher fidelity, and is (in general) smarter. Nonprofits who use ChatGPT for social media graphics, event flyers, or donor communications should find this noticeably better than before.
- Google’s Workspace Intelligence is now live. Google quietly rolled this out in late April — it’s a major update to how Gemini works inside Google Workspace. If your organization uses Gmail, Drive, Calendar, and Docs, Gemini now has access to all of that data by default to answer your questions and help you draft documents. For example, you can ask Gemini to “summarize the meeting notes about Project X from Drive” and it will actually find and use them. As an admin, you have controls over which data sources it can access — worth checking your settings if you have a Google Workspace account for your org.
- Claude for Small Business launched. This is a package of connectors and workflows that embeds Claude into tools like QuickBooks, PayPal, HubSpot, Google Workspace, and Microsoft 365 (Word, Excel, Powerpoint, etc). It’s aimed at small organizations like nonprofits that want AI baked into the tools they already use rather than as a separate chat window. My take on this is that Anthropic is seeing Microsoft get traction with Copilot’s integration, and doesn’t want to get left behind here.
Tip of the Month
Using AI for deep research is incredibly powerful, and something I talk about during our AI In Depth class. But these queries can take a long time to run, and use significant compute resources – which can be bad for both your AI budget, and the environment. So crafting a good prompt is more important that with your typical AI interaction. My approach is to use AI (e.g. via Claude Artifacts) to suggest a good prompt, which I then review/edit before using with deep research.
For example, here’s my interaction with Claude’s Artifacts system to craft the prompt I needed to research newsletters I might want to use to tell nonprofits about our training classes. Note that Claude uses a Q&A approach to create artifacts (the prompt, in my case), and thus all of the questions are auto-generated by Claude.
Q: What would you like to create?
A: A Gemini deep research prompt
Q: What’s the intended output or goal?
A: Structured report
Q: Who is the audience for the research results?
A: Nonprofit staff
Q: What’s the topic or subject the deep research prompt should focus on?
A: Finding newsletters for nonprofits where it would be suitable to include information about the monthly online AI training and free meetup that we provide.
Q: What’s the geographic scope for the newsletters?
A: National (US)
Q: What types of nonprofits should the newsletters serve?
A: Nonprofit sector broadly
Claude then came up with its suggested prompt:
Goal: Find US-based newsletters that reach nonprofit staff broadly (not cause-specific) where AI for Community could place a mention, article, or listing about its monthly online AI training and free meetup. <snip>
I reviewed the prompt, and asked Claude to make a change:
Please have the report also include social media groups/communities, and include information on whether it’s free or paid placement.
Claude then generated a really good prompt, which I used in Gemini to generate a very useful report and plan. One cool thing is that Claude included these tips:
This prompt is ready to paste directly into Gemini’s deep research mode. A few tips for getting the best results:
The Section 3 summary is where you’ll get the most actionable guidance — that’s a good place to start reading.
Use Gemini 3 Pro with Deep Research enabled — this prompt is designed for a multi-step research run, not a single-turn chat response.
Expect 10–20 minutes for Gemini to complete the full research pass.
Follow up after the initial report by asking Gemini to dig deeper on any specific outlet (e.g., “Research the submission process for Nonprofit Tech for Good’s newsletter in more detail”).
Final Thoughts
The AI Backlash is real, I see it regularly when talking to friends and people in my community about AI. And there are many issues with the rapid deployment of AI into so many aspects of our daily life – it feels like we’re not being given much say in how this major transformation of work and (eventually) society will play out.
And one major source of significant angst is the issue of job loss.
Though I just listened to an interview with the CEO of a cybersecurity company. He said that every manager in his company has a list of backlog projects “as long as their arm”. With AI helping them do more, the imperative is to burn through these pending projects, not reduce headcount. Yes, someone who is effective with AI will now likely get a job over someone who isn’t, but that’s a different kind of impact.
And I have yet to meet a nonprofit where they say “You know, we’re doing everything that needs to be done, in fact we should cut headcount because we’ve got people just sitting around doing nothing”.