zerakith

joined 2 years ago
[–] zerakith@lemmy.ml 4 points 8 months ago

They did and I'm not against it in principle but it did read a bit like I understand this better than neuroscientists because its just like computer systems and my understanding from neuroscientists is that these sub systems they outline in papers are highly contested and more like how you would model their function rather than actually how they work.

The one thing they seem to say consistently is that the analogy to computer systems work is flawed and the brain works very differently but it's very far from anything I understand so take that with a pinch of salt.

[–] zerakith@lemmy.ml 16 points 8 months ago (2 children)
  • One person who has no background in neuroscience reckoning of how the brain works.
[–] zerakith@lemmy.ml 2 points 8 months ago

Yes its clear that the path of throwing more and more resource at LLMS to improve quality has been a lazy growth focused approach that we could do better if we actually try a design focussed approach.

For me though it comes back to the fact we are facing a polycrisis and most of our resource should be focused on looking for solutions to that and I'm not sure what problem* this technology solves yet alone what problem relating to the polycrisis.

*I realise what they are designed to solve is a capitalist problem. How can we avoid paying staff for service and creative type jobs to increase profit.

[–] zerakith@lemmy.ml 1 points 8 months ago (2 children)

I agree but there's a lot of detail about what activities a lower energy society precludes and my point is that energy intensive "AI" (mostly thinking about LLMs rather than targets applications of ML) probably aren't part of it.

[–] zerakith@lemmy.ml 3 points 8 months ago

Yes that's fair. I guess my comment wasn't a direct response to yours other than it made me think this desire that all the difficult issues (like bias) just disappear if you remove all the humans from the process* is flawed and any anticapitalist society should really start from that understanding. One that understands that conflict will emerge and pro-social "convivial" systems and structures need to emerge to handle them.

*You are right to point out that the "AI" we are talking about is statistical models built from humans that includes bias where as the hype is that we have Data from Star Trek and therefore these systems hide the human inputs but don't remove them.

[–] zerakith@lemmy.ml -3 points 8 months ago (3 children)

To be honest I'm tempted to say that desire to remove humans from the production of society is a fundamentally capitalist one.

[–] zerakith@lemmy.ml 5 points 8 months ago (4 children)

I get the sentiment and I wish it were true.

Some of the issues stem from material and energy limitations regardless of human organisation structures. Fossil Fuels are stored sunlight over a long period of time that means that burning them has a high yield and that's given us a very high EROI society (one where there's an abundance of energy for purposes that aren't basic functioning).

I recommend reading The Collapse of Complex Societies by Tainter who discussing the energy limitations of society. Its before our understanding of energy limitations of technology and he's by no means a leftist but it is still a good introductory text to it.

[–] zerakith@lemmy.ml 4 points 8 months ago

And none of the issues are helped by a further moving target by pursuing something that pushes our energy usage even higher like some forms of "AI" that produce very little meaningful outside of capitalism anyway.

[–] zerakith@lemmy.ml 3 points 8 months ago (1 children)

Let's say its true that doing that would stop the problem getting worse (e.g. no more emissions after 5 years)*.

We still have the legacy issues to deal with and I need anticaps who are thinking seriously about what can replace capitalism to take seriously how dependent we are on natural systems that are very close to collapse. We are already passed the point where just stopping the harm is job done. The climate is not the one we have evolved and developed civilisation under its far less stable.

  • There are material and energy constraints that aren't instantly solvable and electricity production is far from the only cause of climate harm (land use and manufacturing) and some of those have major question marks remaining as to how they can be removed or electrified.
[–] zerakith@lemmy.ml 6 points 8 months ago* (last edited 8 months ago) (32 children)

They don't disappear if capitalism disappears. I agree with you capitalism needs to end in order to deal with them but there are hard issues that we have to deal with even with capitalism gone.

Even if the causes ceased we would still be left with residual emissions and degraded natural systems to try and deal with and a lower EROI society to do it.

[–] zerakith@lemmy.ml 43 points 8 months ago (69 children)

It doesn't solve the energy and emissions crisis we are facing but sure.

[–] zerakith@lemmy.ml 7 points 8 months ago

https://en.m.wikipedia.org/wiki/Clarke's_three_laws

The third law is "Any sufficiently advanced technology is indistinguishable from magic."

I merely meant that the beauty of mathematics and natural science was a form of magic.

 

Since Reddit is now explicitly planning to sell user generated content for AI training. It got me thinking about Lemmy.

What license are posts and comments assumed to be under on this instance? Is there an overarching lemmy policy (there doesn't seem to be)?

Is it down to the user to specify, if so how?

Are there any downsides with adopting a Creative Commons or other copyleft license?

 

I'm in a bit of a productivity rut and whilst I suspect the issue is mainly between the keyboard and chair I'm also interested in what (FOSS) tools there are that people find effective.

One of my issues at the moment is cross managing different workstreams particularly with personal projects which are more in the "if I have time category".

I'm interested in anything that helps manage time or limit distractions or anything that makes it easier to keep track of progress/next steps for project when there may be a bit of a time gap between.

 

I've been playing some of the more recent adventure games and feel like the quality of the puzzles has gone down. It often seems a bit like use multitool on object to solve every puzzle. Equally, I can think many older games where the puzzle was so illogical it broke the gameplay and felt jarring to me.

So what makes a good puzzle? What are you most satisfying puzzles ever? What about your least favourite?

view more: next ›