A sepia tone of what seems to a young Black girl with her eyes covered by a black rectangle (censor style) with white letters reading "Don't Believe The Hype (remix)"

Beware the AI Hype Cycle

In the months I been seeking my next job in nonprofit tech, I’ve learned that the sector is quite invested in using AI right now. There are so many important applications of LLMs and machine learning, in fact we’ve all been using it for years to do things like sorting our inboxes, analyzing complex data, finding the perfect digital asset for a specific purpose, and showing the perfect next video for our interests. I don’t want to give those things up. (OK, maybe that last one sometimes.)

Frankly, I worry that seeming critical of AI will negatively impact my employment prospects, but there’s a reason that I have a reputation as a trustworthy and ethical tech leader. And I hope that there are still organizations that would rather have an honest technologist instead of one that never challenges them. There are real benefits but we have to tell the truth about the fact that there are also real harms to our organizations, our communities, and our planet from reliance on AI.

I suspect the current normalization of generative AI as simply part of how we all work now is much more based in marketing hype than practical nonprofit sector experience. I learned about the AI hype cycle from researchers like Timnit Gebru and Emily Bender. I’ve been following them since 2021 when Gebru and others were fired from Google’s AI ethics team when their bosses objected to their groundbreaking co-authored paper on LLMs called “On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? 🦜“.

The adoption that big tech wants us to believe is “inevitable” is much more driven by the fact that companies are cramming it in to every piece of software whether users want them to not. Business experts have made it clear that Big Tech is over leveraged in the AI sector. If we don’t use it enough, they don’t make the profit they need. We don’t need AI to survive as much as it needs us to live.

In their latest article, Bender and Gebru once again bring us back down to earth:

The AI industry has even suggested that popular, bipartisan anti-data-center activism is a “distraction” from attempts to regulate the impending, scary, “superhuman” machines these companies are building. According to the AI industry, we should be more worried about a fictional machine god than about the climate catastrophe that these data centers exacerbate, the asthma suffered by those living near them, the rising electricity bills of the public subsidizing them, or the water that is redirected to cooling them.

– Don’t be fooled by this summer of AI hype, 9/22/26

As nonprofits and community-oriented people continue to learn and experiment with AI, it’s essential that we stay focused on the real requirements of our missions and the real consequences of our choices. As Flava Flav taught us, don’t believe the hype!

A closeup of art that has tiny mirrors with circles cut out woven into a chainlink fence

A human starts the loop, a human can stop it

In the social good side of the tech world (hashtag nptech), I hear people and institutions talk about the necessity of having “a human in the loop” when working with AI. This usually means that a responsible person reviews output from an LLM before it gets out in the wild. This is a low standard, but it’s not wrong. I hope to see more discussion of what having a human in the loop means for agentic AI, which is generally sold to us as a way to reduce human labor (for better or worse).

But let’s be serious. Humans need to be more than “in the loop.” Humans need to be held accountable for the results of their actions. Software does not build itself. Agents do not tell themselves when to run. When we talk about these tools as if they “went rogue” and “hacked into” things, we are removing the humans that built the tool, configured the sandbox, and gave the instructions.

AI is not people (nor are corporations but that’s another rant). As written in a 1979 IBM training manual, “A computer can never be held accountable, therefore a computer must never make a management decision.”

Software can’t be bound by the law, but its creators can and they must.

AI Framework for an Equitable World NTEN

A much-needed AI framework for a more equitable world

Since losing my job in a round of budget-cutting layoffs at MomsRising (where I spent 10 years building web sites, IT systems, security protocols, and trainings) I have been seeking a new opportunity to help a progressive organization or campaign to more effectively use technology in support of grassroots organizing for political change. One of the things I have found in every CTO and IT Director position I see is that everyone wants to use AI now, but not everyone is sure why or how.

I am so grateful to have received a scholarship to enroll in classes to earn NTEN’s AI for Nonprofits Professional Certificate. I have been following the discussion around AI safety and ethics for many years, including as MomsRising’s Senior Director of Internal Technology and Security. This certificate program is giving me more practical tools for how to create and guide AI applications that can serve movement organizing while still staying true to our progressive values like justice, equality, and mutual care.

This week, I took a class created by Amy Sample Ward and Tristan Penn about the AI Framework for an Equitable World that was created in a community process led by NTEN in 2023. I greatly appreciate this approach which shows how assessment and impact are essential, on-going elements of any AI intervention. Below is a diagram of the framework. This is exactly the type of useful resource that I and any nonprofit IT leader can use to develop tools and processes that are both more effective and less harmful. I look forward to putting this into use soon!

AI Framework for an Equitable World. 

A diagram with 3 connected loops. 

Assessment: Input, outcomes, accountability 

Impact: Individual, organization, system 

Intervention: Design, develop, deploy

Learn more and download the full report at https://www.nten.org/posts/publication/artificial-intelligence-framework-for-an-equitable-world

Drupal 8

Getting started with Drupal

Last week I attended the Nonprofit Technology Conference for the first time since 2011. I was very impressed with the organization and the the content of the conference, it was great to see how the community has evolved over the years.

Along with Tim Nafziger, I co-facilitated a discussion session for members of the NTEN/Nonprofit Drupal community. We had an excellent group with a wide range of people participating, from newbies to agency owners. Our two main topics were how to get started learning to use Drupal, and what the future of Drupal holds. The latter topic is hard to summarize, but suffice to say there are still a wide range of opinions about Drupal 8. Even more than 2 years after it’s release, there are still 4 times as many sites using Drupal 7 as Drupal 8.

As for learning Drupal, we had many great suggestions. One person shared this cartoon illustrating that rather than the typical “learning curve,” figuring Drupal out is more like climbing a cliff. But when you get to the top you can really kick butt.

The Drupal Learning Cliff

 

So, here are our suggestions for scaling that cliff!

People with red hearts making the shape of a heart. Illustration by Keith Haring.

I (heart) unconferences

As I mentioned in my last blog post, HASTAC is hosting and facilitating the first-ever NCTech4Good Unconference on April 16th. I’ve found that there is a lot of (understandable) confusion about what an “unconference” (or “Barcamp”) is, especially among those who have not experienced them – but sometimes even with those who have.

The history of this idea is less important than how it’s done, but it’s quite interesting and worth mentioning here. This idea was first conceived in the 1980’s as Open Space Technology – a way for participants to organize and conduct their own conferences. The first BarCamp (essentially an unconference for geeks) was held in 2005 in response to the elite FOOCamp  that was hosted by tech media mogul Tim O’Reilly. I believe BarCamps were also inspired by the BloggerCons of the early 2000’s which aimed to bring a blog-like dialog into real-life meetings. So that’s enough history, see this Wikipedia entry for lots more fascinating background: http://en.wikipedia.org/wiki/Open_Space_Technology .

There are a few specific things that I think are key to a successful unconference:

  • There is no agenda before the meeting, but there is a schedule. Some unconferences do kick off with a speaker or panel to spark discussion. It is important to begin the day with someone who can clearly explain the process and lay out the goal of the gathering.
  • Broad participation is key. After kicking things off, anyone who would like to hold a session stands in front of the room and gives a very short “pitch” for their topic. These can be as formal as presentations or as informal as a conversation. The proposer does not need to be an expert on the subject, but has to be willing to facilitate the conversation if no-one else wants to. It’s not uncommon for half the people in the room to offer some kind of session.
  • All of these session ideas are written on cards and posted on The Grid. This is a chart showing your available meeting spaces on one axis and time slots on the other. After all the pitches are posted (or while they are being made) the participants should be prompted to show a general sense of interest (applause, dots, whatever) to help indicate which sessions need bigger or smaller rooms. Then everyone who cares to can have at The Grid – moving and combining (or separating) sessions until everyone’s varying needs are met.
  • In the sessions, The Law of Two Feet is in effect. That is “If at any time you find yourself in any situation where you are neither learning nor contributing: use your two feet and go someplace else.”
  • Another principle is to strongly encourage public documentation of the event. This can be done via blogs, wikis, Twitter, Flickr, etc. This varies depending on the group, but it’s very common that an unconference will have already posted a wiki to collect ideas before the event, so this is a great place to post notes from the sessions.
  • Oh, make sure to provide lots of food and coffee, and have fun!

Registration for the NCTech4Good Conference and Unconference has been held open until Friday April 8, please go sign up now at http://nctech4good.org .

If you haven’t heard enough about unconferences, I strongly recommend the blog of facilitator Kaliya Hamlin. Start here: http://www.unconference.net/unconferencing-how-to-prepare-to-attend-an-unconference/ .

I also recently discovered a wonderful article about teachers using unconferences: http://plpnetwork.com/2011/03/07/unconference-revolutionary-professional-learning/ .

(Also posted at hastac.org and nctech4good.org.)