jedharris.bsky.social
Infovore, devoted grandparent, practical cook, housing remodeler and maintainer, walkable community master developer
58 posts
631 followers
2,920 following
Discussion Master
comment in response to
post
This is great! BlueSky needs more research discovery tools
comment in response to
post
Are you serious? As long as we have Wikipedia, do we need doctors? Will you be calling up your literature professor friends to talk over poems whenever you have questions?
We have *both* AI *and* people. AI can help us by complementing people and making us all more effective.
comment in response to
post
We're also empowered to have this conversation.
Sounds like on the whole you are not a fan of giving people more power.
comment in response to
post
Do you trust Goldman Sachs?
comment in response to
post
Great to know that from your perspective we have all the understanding of diseases and inflation that we want or need. Unfortunately that isn't how things look from where I sit.
Should we cancel all the literature classes that study poetry?
comment in response to
post
Do you agree that if / when AI empowers individuals, that is democratizing?
comment in response to
post
The newest DeepSeek model matches the best previous models but is 45X cheaper to train.
I am running a version of this model in my home computer without any special hardware or high power consumption.
Any tech is less efficient at the beginning.
comment in response to
post
Individuals can also use the tech to help them understand poems, or math puzzles, or graphs of diseases or inflation.
As the tech becomes widely, cheaply available it *can* empower people. Should we trust them to use it to do good things?
comment in response to
post
How can we tell if AI will do more to empower individuals trying to do good things?
Open source models let enormous numbers of people use AI to accomplish their goals. I believe that most people are good.
comment in response to
post
AI models are rapidly getting cheaper to run, and open source ones are rapidly catching up to proprietary ones in ability. This democratizes access.
People find that AI empowers them, as individuals, to accomplish things they couldn't do otherwise.
These are current facts.
comment in response to
post
Recent "reasoning" models have more ability to be self-critical and catch and fix their mistakes. The capitalist imperative will push the tech toward correctness and more creative solutions because that will be worth more money.
So maybe these problems are growing pains?
comment in response to
post
How will the oligarchs control the DeepSeek models? Or the Llama family? Or the Qwen family?
Worrying about the oligarchs is important. But we must not think of them as having magical powers.
comment in response to
post
Absolutely yes! Surprise are a big part of the package. Right now Open AI et al are very surprised at how good open source models have gotten.
Any given technology reduces the *cost* of some activities. Then *people* decide what they want to do with it.
comment in response to
post
The compute will never be free but it is getting much cheaper. I'm running one of the newest models on my home machine now (it is isn't a special machine). People will soon be able to run open source models on their phones, customize them, etc.
comment in response to
post
Maybe you are referring to the algorithms used by e.g. Facebook to show users content? Those do apparently try to "maximize engagement", promoting addiction. However they are not large language models and don't have at all the same design or capability. Plus they are not open source.
comment in response to
post
So partly this is a theological argument?
I would very much like to see your analysis of how the design of these AI models is targeted to get users addicted. I have seen a lot of discussions (pro and con) of the designs but have never seen how the design is set up to achieve this.
comment in response to
post
are bookstore folks fascist when they perform the same office?
comment in response to
post
Thanks! I'll look at that. Didn't mean to imply you are a library, just that you have researched the topic. I have no idea how to trace your statements back to your sources.
comment in response to
post
I also need more info about textile quality and cost across the transition (artisanal to industrial). I thought the cost difference was huge (maybe a factor of 100?) but I don't really know. I don't know anything about quality; I'd guess that both artisanal and industrial quality varied widely.
comment in response to
post
Interesting claims about both textiles and AI. I don't have the historical background to flesh out your textile claims and would appreciate references to good quality historical studies.
comment in response to
post
Amen. I have been doing that and didn't see a single unpleasant reply in this thread. Entirely enjoyable.
comment in response to
post
Thanks! I totally agree that using methods without that background knowledge and due diligence is epistemic malpractice. And I feel your point is important and should be more central in our discourse.
Conversely this is a considerably lower bar than complete understanding.
comment in response to
post
Super helpful distinction. Maybe a slight refinement: Does the author know how to get more information about how M works if they need it? And do they know how to detect flaws in their application of M? If not they can't tell if they are using it in ways that will produce reliable information.
comment in response to
post
“The point is that Grimehug’s vocabulary is probably twice the size of yours,” said Learned Edmund. “And I’ve got no idea how to write it down and I don’t think a human could ever really learn to speak it at all.”
comment in response to
post
From "The Wonder Engine": The thing is that in gnolespeech, he’d be saying twenty different words. The vocal component is only part of it. The rest is in the ears and the whiskers and the posture and maybe some other things I don’t know how to ask about. ... we don’t understand all the nuance.
comment in response to
post
Maybe T Kingfisher is exploring this with her Gnoles. They say "Humans can't smell" because our perceptions are so limited. She also has extended discussions of how their language is more complex than humans and largely untranslatable.
comment in response to
post
T Kingfisher in many of her fantasy novels has a species called Gnoles which pity humans because they are so limited in both expression and perception. Typically they say "Humans can't smell" to reference this. She also has extended discussions of how their language is more complex than humans.
comment in response to
post
In math and contracts, words only mean what you define them to mean, no more and no less.
Different in human discourse.
Is this news?
comment in response to
post
Naturally makes me wonder how cognition is possible. Is your answer the analysis in your paper Tractable Embodied Computation Needs Embeddedness <https://www.rivisteweb.it/doi/10.12832/94728>?
comment in response to
post
@skyview.social tree
comment in response to
post
So obviously the two of us (and your students) mean different things. I was just looking for a clear explanation of what I meant and I sent that to you.
Could you send me a copy editor description that matches your understanding?
comment in response to
post
No. I'm trying to find out how you use that word. I thought it meant editing for low level correctness that doesn't change the style or meaning of the text. When you described the effects of copy editing on your student writing, it clearly did change the style at least.
comment in response to
post
Here's a possibly useful discussion of copy editing vs. line editing which I think captures the distinction you'd want your students to make. How to enforce this when using AI isn't all that clear to me but I hear it can be done.
shirleyrashediting.com/blog/what-is...
comment in response to
post
My guess this that your students are using AI to completely rewrite their material and are calling it "copy editing". So one move would be to help them understand how to **just** use the AI for copy editing. (Assuming there a way to get them to understand that.)
comment in response to
post
We may be using words differently. Copy editing should not change the tone or meaning of the text. Line editing and beyond should increasingly shift the tone and meaning.
comment in response to
post
This sounds like something they'll need going forward. Is there any way to help them learn to discriminate? Recognition is always easier for people than production. If they can learn to discriminate they'll have a tool they can use to judge their own production.
comment in response to
post
Re: reading comprehension, one thing my spouse has found very useful is having each student prepare three questions about the reading material for each class. Because they may have to ask and discuss them in class they are less likely to shirk. I'm not sure how this fits into teaching writing.
comment in response to
post
- Use "too much like what an AI would write" (whether it is AI or not) as a grading criterion. Again given that they'll be in an AI permeated environment this will help them to stand out going forward.
comment in response to
post
- Give the students contrasting examples of AI writing (bad) and the same material with human writing (good) and have them write critiques explaining why the AI writing is bad compared to the human writing. (You can start with the human writing and generate AI writing with the same focus.)
comment in response to
post
- Help the students find constructive ways to use AI (with attribution). For example AI is good at copy editing, especially for students with deficient writing skills. They will have access to AI for support for the rest of their lives, so why not learn to make good use of it?
comment in response to
post
- Treat unattributed AI the same as unattributed copying -- i.e. plagiarism. Because that is basically what it is.
- Have students critique each other. They probably can recognize AI writing. Peer pressure is compelling.
comment in response to
post
We need to brainstorm ways to neutralize the bad aspects of AI and maybe even find ways to leverage it to make things better.
I am not a teacher but some things that occur to me:
comment in response to
post
Very understandable. My spouse who teaches American literature has been worried about this, but so far hasn't been hit with any egregious cases.
I think there's a lot of "demand pull" even more than "provider push" -- students have always sought shortcuts.
comment in response to
post
Interesting and surprising to me that essentially none of the top results have a named author -- expert or not.
As soon as I qualify the title with any theme I get results with named authors (and they are much more interesting).
comment in response to
post
Please remember that while the reasons may be obvious to you, many of us haven't heard or seen what you have.
I hope you (and others commenting here) will make a good faith effort to help us see your point.
comment in response to
post
Please explain the damage in some detail, for those of us who are still ignorant. Please give us some examples from Kevin's actual work.
For us to help you stop this, we need to see at least as much depth of justification as Kevin has provided in his thread.
comment in response to
post
Certainly Kekulé's dream that led him to discover the structure of benzene counts.
comment in response to
post
Interpretation is hallucination. Deconstruction is hallucination. Fiction and poetry are hallucination. History in many cases is hallucination.
If you don't agree please draw the line that separates them.
comment in response to
post
The comments on this post are very good sources of accounts to block or mute. Knee-jerk rudeness with no background knowledge (and often no actual comprehension of the thread as written) typically indicates someone with nothing interesting to say. Don't want to have them pollute my feed.
comment in response to
post
Congratulations on surviving and maybe on enlightening yourself (if your family drank the coolaid). Do you have any suggestions about reaching people in that state?