My gosh I'm so glad I found your publication, it makes me so happy to read in the morning (and makes me question my own publication style - for the better of course!)
This is a great perspective on the AI discourse and as someone who worked in corporate productivity for years, I can absolutely back up that it's mostly hype. There are a few actual use cases, but we do have to be very careful about the workslop.
I can't tell you how many times I had to send a follow up email to a "meeting minutes" email that was obvious AI generated and had tasks assigned to people that didn't exist or sentences that made absolutely no sense without the context from what was presented on screen (not captured by the meeting summary agent.)
“the time you save generating content gets eaten by the time everyone else spends wading through your AI-generated content.”
I’ve been trying to understand exactly why I feel MORE busy and overwhelmed than I did, say, a year ago at work. And that quote is exactly it. I work at a tech firm in a customer-facing role, and AI kind of feels like that coworker who doesn’t really do anything but insists on trying to prove its own relevance by pure verbosity and “insights.”
I get AI slop generated about clients before I meet with them (about 5% of it is helpful and true). I get AI slop from clients. They feed my responses into AI to generate an endless stream of follow-up questions that never allows us to “close the loop.”
This captures the part of the AI productivity story that often gets skipped: output got cheaper, while absorption stayed expensive.
The first draft, summary, deck, or code block has become cheap. The scarce resource is the recipient's capacity to decide what deserves trust. That is why so much AI "productivity" feels like a transfer of labour from the sender to the receiver.
One practical test I think organisations need is a proof budget: for every AI-assisted workflow, define which claims require evidence, who owns verification, what risks need a human name attached, and what output should be deleted rather than shipped.
Otherwise the machine keeps the work in the system and makes it harder to locate.
I've been writing on this from a creative industries perspective (where it is perhaps most notable, given the AI companies' strategic focus on generative creative applications for their products): https://joeathome.substack.com/p/peachy
Hi Hannah - This is the piece I've been waiting for someone to write. The problem that I see is that we haven’t launched enough systems designed to help people navigate a dramatically changing labor market.
We need more infrastructure that can help people build lives with dignity, agency, and community at scale because the models that actually answer that question don't look like jobs at all. They look like cooperatives, timebanks, community land trusts, mutual aid networks.
Chicago just embedded cooperative ownership into its city economic development process. The Platform Cooperativism Consortium just trained 200 people from 33 countries on cooperative AI ownership.
These projects are operational. The crisis you're describing is most definitely real — but the alternatives are already running, they just have a discovery problem.
I’m, curious where your Wednesday piece lands on this. Thank you for your perspective!
Don’t think any one of the couple dozens of men or so who actually brought the core technology of AI thought AI would free us. They did because they could, and also because they made a shit ton of money, and also because it stroke their ego. See Geoffrey Hinton as a prime illustration.
I've seen this in a lot of my students who are now burned out and overwhelmed because they are putting in so much time to use LLMs instead of learning the skills themselves. Even something as simple as writing out an email takes them twice as long because they first get it from ChatGPT.
"There’s also another danger that isn’t getting enough attention: It isn’t just whether AI will replace workers. It’s who will control the knowledge that companies capture from their workers." --> I worry about this. I just saw a job posting that entails training AI to the job with your knowledge. Yikes.
I DREAM of the day we get universal basic income!! I've had these thoughts for a while as we've progressed technologically, and I remember reading a quote from someone 100 years ago that they predicted that by now, people wouldn't be working nearly as much. I wish as a society we could get to a place where people could feel financially secure working 20 hours a week at some office job, but then have that extra time for whatever they want - hobbies, art, spending time with family, literally anything that shouldn't be replaced with AI that makes life better!!! It's a dream but I really believe it could happen one day.
On private schools, Avenues is an outlier in that it is for-profit, which means it is constrained from raising money from donations. Most private schools are not for profit and derive substantial income from donations as well as income from their endowments. This enables them to charge less than Avenues and provide about 15% to 20% of tuition in scholarships providing some economic diversity. I don't know what Avenues does in terms of scholarships.
I was Chair of Riverdale for seven years so I'm very familiar with the financial model for not for profit schools.
My gosh I'm so glad I found your publication, it makes me so happy to read in the morning (and makes me question my own publication style - for the better of course!)
This is a great perspective on the AI discourse and as someone who worked in corporate productivity for years, I can absolutely back up that it's mostly hype. There are a few actual use cases, but we do have to be very careful about the workslop.
I can't tell you how many times I had to send a follow up email to a "meeting minutes" email that was obvious AI generated and had tasks assigned to people that didn't exist or sentences that made absolutely no sense without the context from what was presented on screen (not captured by the meeting summary agent.)
I appreciate it! Welcome!!
“the time you save generating content gets eaten by the time everyone else spends wading through your AI-generated content.”
I’ve been trying to understand exactly why I feel MORE busy and overwhelmed than I did, say, a year ago at work. And that quote is exactly it. I work at a tech firm in a customer-facing role, and AI kind of feels like that coworker who doesn’t really do anything but insists on trying to prove its own relevance by pure verbosity and “insights.”
I get AI slop generated about clients before I meet with them (about 5% of it is helpful and true). I get AI slop from clients. They feed my responses into AI to generate an endless stream of follow-up questions that never allows us to “close the loop.”
It’s maddening and I’m exhausted.
This captures the part of the AI productivity story that often gets skipped: output got cheaper, while absorption stayed expensive.
The first draft, summary, deck, or code block has become cheap. The scarce resource is the recipient's capacity to decide what deserves trust. That is why so much AI "productivity" feels like a transfer of labour from the sender to the receiver.
One practical test I think organisations need is a proof budget: for every AI-assisted workflow, define which claims require evidence, who owns verification, what risks need a human name attached, and what output should be deleted rather than shipped.
Otherwise the machine keeps the work in the system and makes it harder to locate.
I've been writing on this from a creative industries perspective (where it is perhaps most notable, given the AI companies' strategic focus on generative creative applications for their products): https://joeathome.substack.com/p/peachy
Hi Hannah - This is the piece I've been waiting for someone to write. The problem that I see is that we haven’t launched enough systems designed to help people navigate a dramatically changing labor market.
We need more infrastructure that can help people build lives with dignity, agency, and community at scale because the models that actually answer that question don't look like jobs at all. They look like cooperatives, timebanks, community land trusts, mutual aid networks.
Chicago just embedded cooperative ownership into its city economic development process. The Platform Cooperativism Consortium just trained 200 people from 33 countries on cooperative AI ownership.
These projects are operational. The crisis you're describing is most definitely real — but the alternatives are already running, they just have a discovery problem.
I’m, curious where your Wednesday piece lands on this. Thank you for your perspective!
Don’t think any one of the couple dozens of men or so who actually brought the core technology of AI thought AI would free us. They did because they could, and also because they made a shit ton of money, and also because it stroke their ego. See Geoffrey Hinton as a prime illustration.
I've seen this in a lot of my students who are now burned out and overwhelmed because they are putting in so much time to use LLMs instead of learning the skills themselves. Even something as simple as writing out an email takes them twice as long because they first get it from ChatGPT.
"There’s also another danger that isn’t getting enough attention: It isn’t just whether AI will replace workers. It’s who will control the knowledge that companies capture from their workers." --> I worry about this. I just saw a job posting that entails training AI to the job with your knowledge. Yikes.
I DREAM of the day we get universal basic income!! I've had these thoughts for a while as we've progressed technologically, and I remember reading a quote from someone 100 years ago that they predicted that by now, people wouldn't be working nearly as much. I wish as a society we could get to a place where people could feel financially secure working 20 hours a week at some office job, but then have that extra time for whatever they want - hobbies, art, spending time with family, literally anything that shouldn't be replaced with AI that makes life better!!! It's a dream but I really believe it could happen one day.
On private schools, Avenues is an outlier in that it is for-profit, which means it is constrained from raising money from donations. Most private schools are not for profit and derive substantial income from donations as well as income from their endowments. This enables them to charge less than Avenues and provide about 15% to 20% of tuition in scholarships providing some economic diversity. I don't know what Avenues does in terms of scholarships.
I was Chair of Riverdale for seven years so I'm very familiar with the financial model for not for profit schools.