How a Video Game Buried in the Desert Explains What AI Is Doing to Nonprofit Training

This is the opening piece in a four-part series on how nonprofit learning got buried under cheap content, and how to find the good stuff again. The three parts after this one look at reviews, online groups, and case studies.

Key takeaways

• In 1983, the video game business flooded its shelves with rushed, broken games, buyers stopped trusting any of it, and the market fell about 97 percent.
• Nonprofit learning is flooding the same way, and AI is the reason. It now costs almost nothing to churn out a course, a webinar, or a toolkit that looks professional and teaches you little.
• Nonprofits get hit hardest because training budgets are usually the first thing cut, so free and cheap content is often all that is left.
• The last crash got fixed when someone stood between the buyer and the flood and marked the good stuff. Nonprofit learning needs its own version of that.
• The real shift is from content you passively take in to context you practice, tested by real people who do the work.

In 1983, Atari had a problem. It had made millions of copies of a game based on the movie E.T. the Extra-Terrestrial. The game was rushed. One programmer built the whole thing in about five weeks so it would be ready for the holidays, and it barely worked. Players kept falling into pits they could not climb out of. People bought it, hated it, and returned it in huge numbers.

Atari could not resell the extra copies and could not dump them without wrecking its own prices. So one night it loaded fourteen trucks with unsold games, drove them to a landfill in Alamogordo, New Mexico, and buried them in the desert under a layer of concrete. For years people thought that was a myth. Then in 2014, a crew dug the site up and found about 1,300 of the cartridges, E.T. included. The story was true.

That buried game is a good way to understand what is happening to nonprofit training right now.

Let's step back and look at the big picture. The E.T. game did not flop all on its own. It was part of a flood. Anyone could make a cartridge for the Atari, so the shelves filled up with rushed, broken titles. Buyers could not tell the good ones from the bad ones, so they stopped trusting all of it. In about two years, the home video game business in North America fell from roughly $3.2 billion to around $100 million. That is a drop of about 97 percent. All that volume did not build the market. It helped bury it.

Nonprofit learning is filling up the same way, and the reason is new. Until recently, making a real training course or a useful toolkit took time and know-how. Now AI can produce all of it in minutes, for almost nothing. So the field is getting flooded with courses, slide decks, checklists, and "ultimate guides" that look polished and say very little. A lot of it just reworks public tax filings or repeats advice you already know. It is the E.T. cartridge in a shinier box, and there is a mountain of it.

I have started calling this the E.T.-ification of nonprofit learning. Cheap, hyped, look-alike content piling up faster than anyone can sort it, until people stop trusting any of it. The game industry had a name for the junk that caused its crash. They called it shovelware, because it was shoveled onto shelves to fill space. We now have shovelware for the social sector, and your inbox probably proves it.

I am not guessing at this. I came into a nonprofit software company as its seventh employee, and over the years that followed I personally helped more than eight hundred nonprofits invest into that platform, most of them built a lot like yours. I also built learning spaces for the field, including a yearly online conference that drew thousands of people at once. That work gave me a close look at what nonprofit teams need, how they go searching for it, and how easy it is to sell them something that only looks like help.

Why nonprofits get hit hardest

There is a reason our field is such an easy target. For years, nonprofits have been told to keep "overhead" low so donors will see them as lean and worthy. Training and staff development count as overhead, so they are usually the first things cut. People call this the starvation cycle, and the numbers are rough.

One long study found that foundations put a little over 1 percent of their grant money, on average, toward building the skills of the people they fund. In another look at grant applicants, 77 percent of nonprofits spent less than 1 percent of their budget on staff training. Compare that with the business world, where companies spend somewhere between $846 and $1,283 per employee each year on formal learning. Plenty of nonprofits spend close to nothing. It is no surprise that almost 90 percent of nonprofit leaders say they worry about burnout.

So when your training budget is small or gone, free and cheap content is what is left. And when you cannot tell whether a free course will help or just eat an afternoon, you are taking a gamble every time. The economist George Akerlof described this trap back in 1970. When buyers cannot judge quality before they buy, the cheap junk crowds out the good work, because good work costs more to make and cannot compete on price. He called it a market for lemons. Nonprofit learning has become one.

Here is the same pattern, then and now.

How the last crash got fixed

The video game business did not stay dead. It came back, and the way it came back is the lesson I'm hoping we learn. In 1985, Nintendo launched a new console with one firm rule. Nothing ran on it unless Nintendo approved it. They put a matching chip inside each console and each game, and if the two did not match, the game would not start. They also put a gold sticker on approved boxes, the Official Nintendo Seal of Quality, that told buyers this one was tested and it works.

It sounds simple, and it worked. Buyers learned they could trust the seal, so they started spending again. The lesson is that trust came back only when someone stood between the buyer and the flood and said, plainly, this one is good and that one is not.

What a seal of quality looks like for us

Nonprofit learning needs its own version of that seal. It would not be a gold sticker. It would be a simple way to know that the thing in front of you was tested by real people and actually helps.

Free courses are not the only thing people reach for. A lot of us gave up on courses and turned to each other instead, in Facebook groups and online forums. That instinct is a good one, and community is the whole point of this series. Yet a crowded forum has its own problems. The loudest voice tends to win, and the wisest one gets lost. Advice that worked for a big-city hospital gets handed to a tiny rural food bank, where it falls apart. And few leaders will admit their real problems in public, so the talk stays shallow. We will get into all of that in a later part.

The kind of learning I believe in is hands-on, the kind you practice instead of sit through. Studies of adult learners are blunt about this. As little as 5 percent of what people take in from a lecture or a webinar shows up later in how they actually work. You remember what you do. That is the idea behind the Generosity Roundtable, the card-based game my team and I built and tested with more than three hundred nonprofits. Instead of watching a slideshow about donors, a board and staff sit around a table, play through real situations, and talk honestly about money, trust, and old habits before those things blow up for real. It has a built-in safety valve too, a card anyone can play to pause the game if a conversation gets too heavy. Think of it as a seal of quality you can feel, because you lived it for an hour.

What is coming in this series

Over the next few weeks I will take the same lens to the three places nonprofits go looking for answers, and show you how to tell the real thing from the flood.

In this series

• This piece lays out the big picture on why cheap content is burying good learning, and what a fix looks like.
• Part 1 shows how to read reviews and ratings, and why the middle reviews tell the truth.
• Part 2 looks at where peer advice in Facebook groups and forums helps, and where it falls short.
• Part 3 breaks down how success stories and case studies get manufactured, and how to check them.

None of this means the tools are bad. AI can be a real help when you point it at your own problem and make it do the reading for you, pulling the useful parts out of a long thread or a vendor's pitch. The trouble starts when you let the flood do your thinking for you.

So before you sign up for the next free webinar or download the next ultimate guide, ask one plain question. Who tested this, and did it change how they actually work? If nobody can answer, you are probably looking at a cartridge headed for the desert.

What is one training or tool you keep meaning to try, and who could you ask whether it is worth your afternoon?