Author | Lightnews

Rune

@runefar.bsky.social

440 followers 2.4K following 5.1K posts

He/Him/They(gender apathetic cis-male) Half Norwegian American whose pleasures in life include exploring different topics and considering how to build solutions to different issues. Has a background and education in technology, anthropology and psychology

Posts Media Videos Starter Packs

Pinned

Rune @runefar.bsky.social · Feb 16

I thought I would throw this out there. If anyone wants to ever ask me anything about my experience having a disability, my experinces as someone coming from a mixed cultural and immigrant background, or anything about how that affected things like my education or perspectives; please feel free too

2 2 14

Rune @runefar.bsky.social · 3m

Though over following a standard neural network copy for copy could as is pointed out in the paper suggest some weirdness that there is a difference in memory whatsoever(though I also think how people think about this when folk discussing it would probabily lead to a similar problem of assumed path)

Rune @runefar.bsky.social · 6m

Equally purely mechanistic system would of course not explain why some events around brain damage affect people's memory . Yet when paired it can lead to interesting explanations such as giving understanding to why HM was unable to form new few memories but still retained his existing memories

Rune @runefar.bsky.social · 15m

Like most people models of how the brain work would not be able to explain journals.physiology.org/doi/full/10....

Neural Mechanisms of Spatial Selective Attention in Areas V1, V2, and V4 of Macaque Visual Cortex | Journal of Neurophysiology | American Physiological Society

Luck, Steven J., Leonardo Chelazzi, Steven A. Hillyard, and Robert Desimone. Neural mechanisms of spatial selective attention in areas V1, V2, and V4 of macaque visual cortex. J. Neurophysiol. 77: 24–...

journals.physiology.org

Rune @runefar.bsky.social · 22m

but that also shows how we need to be better as a whole with clarifying what parts are what

Rune @runefar.bsky.social · 23m

their conscious experience of learning without thinking of the unconscious side. This means that even things like the distinctio between appreciative and associative aspects whn it comes to visual agnosia dont fully connect with people folk model

Rune @runefar.bsky.social · 25m

In fact, the paper you linked alonso and mondragon basically end up confirming this aspect while still showing how other aspects of it is are built on convenience. Many people in conversation deny even basic mechanism of feedback itself as being involved in learning because they focus solely on

Rune @runefar.bsky.social · 32m

or more generally parallel distributing procssing as a whole link.springer.com/rwe/10.1007/...

Parallel Distributed Processing

link.springer.com

Rune @runefar.bsky.social · 47m

Or recurrent processing theory?

Rune @runefar.bsky.social · 49m

What is your take on global workspace theory if I may ask?

Rune @runefar.bsky.social · 49m

My experience discussing this has been that people are in fact also denying elements that we do know in fact are similar to neural networks not the sum idea that they are 1:1 but it is specific to conversation

Rune @runefar.bsky.social · 52m

Here is the issue: it is a bit of a yes and no situation. Because it doesn't resemble it in totality, but it does often resemble it in relation to what people are specifically referring to which often are elements similar hebbian learning on a general level.

Rune @runefar.bsky.social · 3h

When I have posted it elsewhere it is certainly getting feedback. Mainly of the "this is a nightmare generator" kind. Though these ethical considerations are important, I also see a lot to build on for disabled individuals like me and understanding the mind in general.

Rune @runefar.bsky.social · 3h

And people try to use any precident changes as they can

Rune @runefar.bsky.social · 3h

I agree with the decision but I admit this one is one where I am slightly more mixed on a personsal level compared to say putting the old testament on the wall and preventing other religious figures as it looks slightly more private. Decision upholds president though

1 1

Rune @runefar.bsky.social · 3h

Also we are getting closer to that blue neuron lol xd www.nature.com/articles/d41...

‘Mind-captioning’ AI decodes brain activity to turn thoughts into text

A non-invasive imaging technique can translate scenes in your head into sentences. It could help to reveal how the brain interprets the world.

www.nature.com

Rune @runefar.bsky.social · 4h

I was a bit worried this was gonna be a dismissal of men's experiences when I saw this but it does highlight the actual important point. There is an issue but isnt LGBTQ or gender equality, it is that young men feel like they aren't seen as part of the cooallition or recieve support

Charlotte Clymer @charlotteclymer.bsky.social · 6h

"All of these candidates who won last night didn’t shy away from talking about gender equality and LGBTQ rights. They didn’t buy the theory that doing so would alienate young men. And they were right. Stick that in your pipe and smoke it, Mr. Carville."

charlotteclymer.substack.com/p/what-young...

What Young Men Problem?

Democrats won clear majorities of this constituency.

charlotteclymer.substack.com

1 1 1

Rune @runefar.bsky.social · 4h

I think advertisements using AI are one example that really pulls out these individuals. It can be interesting to observe as often what they criticize is stuff observable in the non ai ads because they were the companies style

1 1

Rune @runefar.bsky.social · 5h

Though there will be some ethical ramifications for this, this seems like a amazing development in our ability to peer into both our mind and develop disability technology

Brendan Maher @bmaher.bsky.social · 5h

'Mind captions' sound pretty cool. This work covered by my colleague promises to create text descriptions of the things you see by decoding your brain activity while viewing them. In testing it also worked when people recalled watching videos.

www.nature.com/articles/d41...

‘Mind-captioning’ AI decodes brain activity to turn thoughts into text

A non-invasive imaging technique can translate scenes in your head into sentences. It could help to reveal how the brain interprets the world.

www.nature.com

Rune @runefar.bsky.social · 5h

This is why like I mentioned things like the internet archive or libraries have gotten into trouble

Rune @runefar.bsky.social · 5h

Actually rereading this, this isn't even overfitting. This is just it being in the training set at all. Similar to a work being in the library of someone who is making art. This gets complicated by what permission you give at the point of sale(something like libgen would count as piracy forexample)

Rune @runefar.bsky.social · 5h

But I'll end the conversation like you said. Please watch the videos and articles though to learn more about thw mechanisms of it.

Rune @runefar.bsky.social · 5h

Though copyright in general always relies on some ambiguity between cases we should not want fair usage and transformativeness gone. AI regulation is one thing but it is not good for leftist to support this and we can see how other countries are reacting in a more focused manner yet supportive

Rune @runefar.bsky.social · 5h

This is why seeing people argue for increased copyright strictness si scary. What people are calling theft would effectively make all t transformative works eroded and create what lessig call a permission culture. This benefits the corporations the most

Rune @runefar.bsky.social · 5h

No I did becausw thar is what the parameters are. They are usage of the facts of the works within a greater work.

Rune @runefar.bsky.social · 5h

That is why I specified output liability from training data because output should absolutely still be liable