Hexorg

joined 2 years ago
MODERATOR OF
[–] Hexorg@beehaw.org 4 points 6 days ago

Its an interesting perspective, except… that’s not how AI works (even if it’s advertised that way). Even the latest approach for ChatGPT is not perfect memory. It’s a glorified search functionality. When you type a prompt the system can choose to search your older chats for related information and pull it into context… what makes that information related is the big question here - it uses an embedding model to index and compare your chats. You can imagine it as a fuzzy paragraph search - not exact paragraphs, but paragraphs that roughly talk about the same topic…

it’s not a guarantee that if you mention not liking sushi in one chat - talking about restaurant of choice will pull in the sushi chat. And even if it does pull that in, the model may choose to ignore that. And even if it doesn’t ignore that - You can choose to ignore that. Of course the article talks about healing so I imagine instead of sushi we’re talking about some trauma…. Ok so you can choose not to reveal details of your trauma to AI(that’s an overall good idea right now anyway). Or you can choose to delete the chat - it won’t index deleted chats.

At the same time - there are just about as many benefits of the model remembering something you didn’t. You can imagine a scenario where you mentioned your friend being mean to you and later they are manipulating you again. Maybe having the model remind you of the last bad encounter is good here? Just remember - AI is a machine and you control both its inputs and what you’re to do with its outputs.

[–] Hexorg@beehaw.org 1 points 3 months ago

Apparently Nintendo switch 2 is using the standard already, so it might go over better than Sony.

[–] Hexorg@beehaw.org 7 points 5 months ago

I think I’m starting to get back into beehaw, so 👋 it’s been a crazy several years…I’ve been diagnosed with clinical depression and now I’m better so that’s good overall

[–] Hexorg@beehaw.org 1 points 5 months ago

It’s an interesting and hard problem. Because most billionaires don’t own billions in cash - they own companies that are worth billions. These companies also don’t have billions of assets - they are valued at billions by investors.

The problem is that musks and bezoses of the world didn’t start with billions - they started with millions and lucked out. So to prevent this from happening you need some system that can fairly catch a moment where a business becomes too big and do something about it.

You can’t really cut the majority owner out, because well they own the company - you can’t just take away what they own. But you can’t really pay them some ceiling cost either - you’ll just end up making someone else a billionaire.

[–] Hexorg@beehaw.org 4 points 5 months ago

I don’t know about more emotional on average, but I can totally see how emotional repression can lead to bigger emotional outbursts.

[–] Hexorg@beehaw.org 2 points 1 year ago

Yup, I’ve been plagued by this bug for a long time. I’m very excited to use this!

[–] Hexorg@beehaw.org 1 points 1 year ago* (last edited 1 year ago) (1 children)

Good point! I wonder if we’re spoiled by computer invention though. Would be interesting to compare preWW2 invention rates and now. I suspect computers just made everything else easier, but now we’re back to hard problems

[–] Hexorg@beehaw.org 5 points 1 year ago (3 children)

To be fair, there’s only been 24 year’s of 21 century. Most things you gave listed happened at the end of the 20th century. But also the question is somewhat self negating - we won’t know what’s the greatest invention until we see it working great, but it takes much more than 24 years to take an invention from concept to consumption. For example computational biology is kicking off. Computer aided dna generation started in the past 24 years. But it’s so new few people think about it. Just like no one thought of internet as the greatest invention in the 70s… it was just too new

[–] Hexorg@beehaw.org 14 points 2 years ago

Alternatives or not, I think it’d be very beneficial to document concept of operation that you want. That way you can either take pieces of these conops and tell lemmy devs what you want, or if you have your own project this will be its conops and you can guide developers towards features you need.

[–] Hexorg@beehaw.org 1 points 2 years ago

That’s because the full version of that mentality is “Tax me less, don’t use my tax money to subsidize someone else, give that money to my company!” Instead

[–] Hexorg@beehaw.org 1 points 2 years ago (1 children)

To be fair even in trek - there’s a world war 3 that’s driven by pure greed before humanity decides it’s enough. And the climax of the greed and that war starts in 2026… so we might be on the course to the utopia … but not before suffering some more.

 

I'm building 90s themed arcade In my shed... but I still want to keep a little workshop area so I'm splitting my shed into 3 rooms including the attic/upstairs.

I've never done construction aside from small things like routing cat6 through the house, so I decided to practice virtually first - I've reconstructed my shed's frame in Blender and added all the lumber that I need to add the second floor. I've also 3d-scanned the current structure and superimposed it in blender so it was a bit easier to see if what I'm doing is sane at all.

Bonus: 3dscan video:

I have a laserdisc collection with a few CRT TVs, Pentium 3 computer with Windows 98, and PlayStation1. I'm also planning on building a few arcade cabinets with emulators.

 

I want to start a discussion of MIT vs GPL and see what you all think

 

I had the weirdest of a problem. Two computers communicating with each other over ping and TFTP works. When I boot one of them into U-boot (a bootloader that supports TFTP boot) it can’t ping not load tftp of the other machine complaining on ARP timeouts.

I swapped with a dumb switch - all works. Everything else (machines, cables) are the same. The managed switch is a Cisco switch and I have a serial console to it, but I’m not familiar with managing those switches - what feature is potentially blocking u-boot's arp packets?

I’ve double checked with tcpdump - the other machine never seer u-boot's arp packets, but does when the same board is booted into Linux. I’ve also checked Cisco's monitor event-trace arp continuous and it didn’t print any packets but it did say link status went from up to down to back up when I rebooted.

Is there some sort of Mac filter on Cisco switches?

20
submitted 2 years ago* (last edited 2 years ago) by Hexorg@beehaw.org to c/chat@beehaw.org
 

Y’all seem like a bunch of friendly folk. I moved to the research triangle, North Carolina during the pandemic and need to rebuild my social circles. Unfortunately I work from home and have a toddler and an infant so meeting new people in real life has been extra challenging. Anyone in the area here? I’m building a 90s themed arcade in my shed, ask me anything!

 

I googled "missing medieval servant" and it came back "page not found."

 

Russian propaganda machine was full force in America during the 2016 election and I’m sure it still is quite active to an extent in the states. What’s the general feel for Russian propaganda in Ukraine? Are there many Russian-supporting Ukrainians? Do many of them change opinions once Russia is in their towns?

Uhh I should note that I’m quite ignorant on the subject and if my question makes you irritated - I’m sorry I don’t mean to start anything.

 

https://github.com/angr/angr Uses a Concolic execution engine where it can switch from running a binary concretely, break, and then define an unknown input and find what should I be to trigger a different breakpoint. - e.g. what should the “password” pointer be pointing to in order to trigger the “you’re in” branch of code.

Note: it still can’t reverse hashes. If you try to reverse md5 using this approach it’ll consume petabytes of RAM.

I think radare2 was looking into integrating with angr but I don’t know the status of the integration.

 

Weka implements a ton of statistics-based ML algorithms as well as some validation tools and graphs. All you need is some data in almost-CSV format and you can run some statistics analysis on it. This isn’t neural networks so you don’t need neither a powerful GPU nor gigabytes of data. Some tutorials online get useful results with 10-20 entries.

view more: next ›