"Who AI benefits and who it immiserates often is based on who gets to decide how it’s used. We know how tech is used on the day to day. We should be at the table as well."
Proud member of local 2627, IT professional employees of the City of New York, DC 37. I was wondering if you knew anything about what I guess I would call 'ethical' LLM models, that is, models that don't use copyrighted material or rip off artists, etc. There is a model in Switzerland (!) associated with an organization call Public AI, https://chat.publicai.co/, which claims to adhere to strict standards, and they have an 8b and 70b parameter model (just ask the chatbot). I've not been able to find any other models like this, but people should be using them, instead of frontier models, if they exist. Thanks for all your work, Jon
"I think many of us felt in the UC that creeping sense of being left out of decision making in how to implement technology for the public good.": My own observations and experiences leave me in no doubt that the chancellors and other senior administrators of the UC are, in the main, weasels and money-grubbers.* Many would fit right into corporate America: the continual shafting of the people who do the real work, the PR-speak to deflect and disguise, and the bloated salaries (at taxpayer expense) that still aren't enough for some of them.** So the IT employees absolutely need to be unionized, and they need not to be shy about striking or threatening to strike, just as teaching assistants at Davis did when I was one of them.***
*I have degrees from three branches of the UC (Irvine, Santa Barbara, and Davis), and I was a staff member at another (Berkeley). I was personally acquainted with two chancellors, one of whom went on to be president of the whole UC. I wouldn't have trusted either of them as far as I could throw them. For example, Larry Vanderhoef promoted the obnoxious bullshit that teaching assistants shouldn't be unionized because we weren't actually laborers but "apprentices" (as if apprentices, in trades that have them, weren't unionized).
**For example, Linda Katehi, who succeeded Larry Vanderhoef as chancellor at Davis, was eventually pushed out for a bunch of shit she did, including raking in ample "compensation" for "serving" on the boards of DeVry Education Group, an operator of for-profit "colleges", and John Wiley, a textbook publisher noted for price gouging. Because her $400,000-per-year (plus generous benefits) "compensation" from the UC just wasn't enough for her.
***The regents and executives refused to recognize our union, so we struck. And we won! (It helped that the California legislature, controlled by Democrats, expressed displeasure with the regents and executives. Little as the latter may like it, the UC is a state institution, and the legislature ultimately holds the purse strings.)
Hearing about your ten year old gave me a big lift. I've got a son (who's 25) and he also makes me proud with his (union and non-union) activism...and so many other reasons as well, of course. Great post, BTW. The growing unionisation gives some real hope...
It was unsuccessful for Lucas because management did not back the initiative. Lasting change only happens when employees set up new companies and organisations with better foundational values.
I would apply that principle to higher education too, because while current university employees may gain benefits from organising, the sector as a whole is based on unsustainable levels of student debt and frequently produces graduate under-employment too.
I'd like to think that a better model is possible, in which the university is not merely a credential factory for a post-industrial economy of managerialism, but delivers broader social benefits.
This is not really a union point, it is a systems point. And it is the one almost no one in tech leadership is saying clearly.
What UPTE is doing is not only protecting jobs. It is saying that if you work close to a system, you should have a voice in how that system runs. That is a different point, and it is much harder to brush aside.
The tech industry has taught us to see this kind of idea as being against progress. But it is really one of the oldest ideas in good system design. The people who use and run a system know things the people who designed it often do not. If you ignore that, the system does not become smarter. It becomes more fragile.
The Politico case shows this well. The AI tool made mistakes. It broke style rules. It left out students. That was not because AI cannot work in theory. It was because no one with real editorial judgment was asked before it was used. That is not a technical failure. It is a governance failure.
So good to have a newsletter full of good news like this, thanks Brian! Big Congrats on your son's work, that's some pretty cool stuff. I raise meal worms for my aquaponics system (though no thanks - I'll pass on feeding them styrofoam!)
Proud member of local 2627, IT professional employees of the City of New York, DC 37. I was wondering if you knew anything about what I guess I would call 'ethical' LLM models, that is, models that don't use copyrighted material or rip off artists, etc. There is a model in Switzerland (!) associated with an organization call Public AI, https://chat.publicai.co/, which claims to adhere to strict standards, and they have an 8b and 70b parameter model (just ask the chatbot). I've not been able to find any other models like this, but people should be using them, instead of frontier models, if they exist. Thanks for all your work, Jon
"At the end of the day, who AI benefits and who it immiserates often is based on who gets to decide how it’s used."
Indeed, I think works should a say in AI is used - but perhaps more importantly, whether it is used at all ...
Go gauchos
Your posts have never made me cry before. The last part about your son winning the science award is truly touching.
That’s a very human demonstration and always very welcome. Congrats on both counts.
The headline story is definitely good news.
"I think many of us felt in the UC that creeping sense of being left out of decision making in how to implement technology for the public good.": My own observations and experiences leave me in no doubt that the chancellors and other senior administrators of the UC are, in the main, weasels and money-grubbers.* Many would fit right into corporate America: the continual shafting of the people who do the real work, the PR-speak to deflect and disguise, and the bloated salaries (at taxpayer expense) that still aren't enough for some of them.** So the IT employees absolutely need to be unionized, and they need not to be shy about striking or threatening to strike, just as teaching assistants at Davis did when I was one of them.***
*I have degrees from three branches of the UC (Irvine, Santa Barbara, and Davis), and I was a staff member at another (Berkeley). I was personally acquainted with two chancellors, one of whom went on to be president of the whole UC. I wouldn't have trusted either of them as far as I could throw them. For example, Larry Vanderhoef promoted the obnoxious bullshit that teaching assistants shouldn't be unionized because we weren't actually laborers but "apprentices" (as if apprentices, in trades that have them, weren't unionized).
**For example, Linda Katehi, who succeeded Larry Vanderhoef as chancellor at Davis, was eventually pushed out for a bunch of shit she did, including raking in ample "compensation" for "serving" on the boards of DeVry Education Group, an operator of for-profit "colleges", and John Wiley, a textbook publisher noted for price gouging. Because her $400,000-per-year (plus generous benefits) "compensation" from the UC just wasn't enough for her.
***The regents and executives refused to recognize our union, so we struck. And we won! (It helped that the California legislature, controlled by Democrats, expressed displeasure with the regents and executives. Little as the latter may like it, the UC is a state institution, and the legislature ultimately holds the purse strings.)
Hearing about your ten year old gave me a big lift. I've got a son (who's 25) and he also makes me proud with his (union and non-union) activism...and so many other reasons as well, of course. Great post, BTW. The growing unionisation gives some real hope...
great read. it’s honestly been pretty interesting to watch this all unfold.
A similar effort to produce socially useful technology was made in Britain in the 1970's by workers at Lucas Aerospace, known as the Lucas Plan:
https://lucasplan.org.uk/story-of-the-lucas-plan/
It was unsuccessful for Lucas because management did not back the initiative. Lasting change only happens when employees set up new companies and organisations with better foundational values.
I would apply that principle to higher education too, because while current university employees may gain benefits from organising, the sector as a whole is based on unsustainable levels of student debt and frequently produces graduate under-employment too.
I'd like to think that a better model is possible, in which the university is not merely a credential factory for a post-industrial economy of managerialism, but delivers broader social benefits.
Hell yeah. ✊
This is not really a union point, it is a systems point. And it is the one almost no one in tech leadership is saying clearly.
What UPTE is doing is not only protecting jobs. It is saying that if you work close to a system, you should have a voice in how that system runs. That is a different point, and it is much harder to brush aside.
The tech industry has taught us to see this kind of idea as being against progress. But it is really one of the oldest ideas in good system design. The people who use and run a system know things the people who designed it often do not. If you ignore that, the system does not become smarter. It becomes more fragile.
The Politico case shows this well. The AI tool made mistakes. It broke style rules. It left out students. That was not because AI cannot work in theory. It was because no one with real editorial judgment was asked before it was used. That is not a technical failure. It is a governance failure.
We need more of this thinking, in more places.
So good to have a newsletter full of good news like this, thanks Brian! Big Congrats on your son's work, that's some pretty cool stuff. I raise meal worms for my aquaponics system (though no thanks - I'll pass on feeding them styrofoam!)