@[email protected] I'm sorry about that. I find myself in frustrating arguments about AI as well. That's partly why I post about it here, to refine my views.

I feel like these AI conversations can end up frustrating for everyone involved partly because there is a host of prejudices and blind spots that come together around this technology. For instance, a whole lot of people, especially Americans, are energy blind--they don't have a very good sense for how much energy it takes to do stuff, where that energy comes from, and what it takes to generate and transmit energy. Some of the more outlandish "proposals" about powering the hyperscale data centers behind generative AI ambitions would have us build more nuclear reactors than there is uranium to fuel, or deploy more batteries than we have minerals to build. Most people aren't informed enough to say "uh, that's physically impossible to do?" and some end up believing these prognostications. I also feel that a whole lot of people aren't really aware of just how much outright theft of material has occurred to create things like ChatGPT. Just straight up taking such an enormous amount of material that creators put so much hard work into is unsustainable. It will have knock-on effects, like people refusing to put their work on the internet for fear of having it stolen, that degrade life for everyone (it's a kind of tragedy of the commons perpetrated by a small number of rapacious actors). Is that the world we want to live in just to have a whiz bang toy that a lot of people say makes their jobs harder?

I feel like it is. It seems to be changing form from "left behind" to "you'll become a failure/obsolete". Certainly the Adobe and Rosetta quotes I cited have the latter, more threatening, tone.

Neither are true, though, not as written nor factually. You won't be "left behind". Instead, bosses/employers/owners will actively and knowingly modify job descriptions, fire or lay people off, etc. to bring about conditions where people they've chosen to exclude are "behind". This isn't gravity, it's human decisionmaking, and the decisions could be made differently. That's why I characterize rhetoric like this as threatening: the powers that be are threatening to take people's livelihoods away and use AI as an excuse, in an attempt to avoid blame for their destructive choices. That's also why I characterize it as coercive control: they are trying to control people's behavior and choices using veiled threats combined with trying to erode people's sense of self worth ("you'll end up a failure, a loser, if you don't do what we want").

I've spoken to quite a few people in and around #tech who've expressed sentiments along the lines of the magic being gone, it not being fun anymore, or generally it changing in negative ways in the last few years especially. I think the expressions by tech leaders of open hostility towards workers and users, as well as the open embrace of #eugenics -- which is fundamentally a hateful ideology -- by so many tech leaders go partway towards explaining this shift. My read is that folks are sensing that the sector has become hostile to most people, and the experience of that sensation is that the magic is gone, or the products aren't interesting, or it's not fun anymore, etc. I would say that something more sinister is going on.

Relentless repetition of dehumanizing language is a pillar of coercive control, which is really what generative AI has been and will continue to be about.

We don't need to live in a world where things we don't want are regularly forced upon us by people who have more power than we do. We can reject that world even when we have little choice but to navigate it.

#AI #GenAI #GenerativeAI #CoerciveControl

It's interesting how the rhetoric around #AI shifts around and now companies are using phrases like "embrace #AI or face extinction". I'm thinking of Adobe's recent move to force artists to use the AI features in their products, under the threat they are "not going to be successful" if they don't; or Rosetta announcing that linguists need to use Rosetta's AI features or face the "extinction" of the languages they work on.

It's a short step from "extinction"/"unsuccessful" ("low fitness") to "elimination". The latter word is what is meant. The passive voice/inevitability framing purposely obscures the agency of the literal, nameable human beings who are attempting to bring this reality into existence. "Embrace #AI or we will do our best to eliminate your profession, your livelihood, and you" is more precise and brings out the hostility of the threat these corporate statements attempt to hide.

This dehumanizing and ultimately eugenic idea frequently hides in plain sight like this. Sometimes evolutionary or genetic language and metaphors are used. Don't accept it. These folks may try to create this reality but that doesn't mean they'll succeed and it doesn't mean we need to surrender and let them succeed.

#GenerativeAI #GenAI #AI #eugenics

@[email protected] The things you listed are almost all animal. Which is part of being human, but not all there is.

I feel like the owning class has always aimed to dehumanize those of us not in it; they couldn't hold slaves, oppress peasants, conquer or colonize otherwise. That's kind of how that relationship works.

@[email protected] There is a lot you can do about, one class of action being to re-assert your humanity at every opportunity.

For folks keeping score, the rhetoric around #GenerativeAI contains deeply dehumanizing language about everybody who is not wealthy. Microsoft CEO Satya Nadella referring to "virtual employees" (1) and suggesting that #LLM ingestion of copywritten works is no different from humans reading textbooks (2) are recent cases in point, but this kind of language is rampant.

We live in a time where we're constantly bombarded with dehumanizing language. A significant amount of this is coming from "the top". I find it unacceptable, personally.

#AI #GenAI #GenerativeAI #LLM #ClassWar

(1) Referring to computer programs. Which employees are not, unless Nadella intends to start paying tax to the IRS on every piece of software every Microsoft employee uses.
(2) These are fundamentally different processes. The work this suggestion is doing is leading one to believe that human beings creating or reading text are no different from LLMs. That is exceedingly dehumanizing, and is quite alarming coming from someone in Nadella's position.

Got my mail-in ballot. I firmly believe that significant action outside the electoral system is necessary to make progress on the many problems that face us. I think of voting a bit like brushing your teeth, a hygienic action that helps forestall decay.

#USPol #MainePol #MEPol

@[email protected] @[email protected] @[email protected] @[email protected] As someone in the US who's been to Pennsylvania--a key battleground state--recently, I'd say this election is uncomfortably close. I have no special knowledge of the matter, but that's my gut on it.

People forget 2016 way too easily. I was teaching a data science class that year and I made a whole lot of hay out of Nate Silver's projection plots that showed Clinton having a near-100% chance of winning, right up until the day of the election when that projection dramatically flipped to Trump having a near-100% chance of winning. The question to the students being a pedagogical version of "wtaf?".

@[email protected] You are being swindled. I honestly think all of us are better served by putting every ounce of spare energy we have into resisting all this stuff. "Reducing your carbon footprint" is a pointless waste of time. Permanently constraining a billionaire's ability to increase their carbon footprint beyond all reason? Now that would have an impact.

Regarding that last boost: pardon my language, but Eric Schmidt is out of his f*#$@ing mind. I never thought all that highly of him but holy cow, lately he's uncorked the nonsense champagne and it's fizzing everywhere. It's extraordinarily discouraging that people like this have the ear of US government officials.

@[email protected] @[email protected] @[email protected] Back when I was a software developer someone in my group identified a subtle bug in money processing code that in some circumstances would have resulted in a small rounding error. We pored over this awhile and convinced ourselves that we could have siphoned money out of that and, since we controlled all these systems, it could probably gone undetected for quite some time. This company had revenues in the many hundreds of millions so it would have added up to a tidy sum over enough time.

We alerted the management and it was fixed. A few things occur to me:

1. An LLM would never find a bug like that. In fact, it's fairly likely to generate them
2. Software developers who are not deeply attuned to their codebase (say because LLMs generated substantial portions of it) would be unlikely to find such bugs
3. If i were a software developer today and I were required to use LLMs in my work, I would not tell management about bugs like this if I found them. Because they've signalled to me they don't respect what I'm capable of enough to support me in it, so why should I? I'ld save the good energy for my hobbies and phone it in at work as much as I could get away with; it won't matter to the higher ups.