--- Log opened Sun Mar 19 00:00:14 2023
02:02 -!- cthlolo [~lorogue@77.33.23.154.dhcp.fibianet.dk] has joined #hplusroadmap
03:10 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has joined #hplusroadmap
03:13 -!- Guest6537 [~Guest65@85.136.168.47.dyn.user.ono.com] has joined #hplusroadmap
03:18 -!- Guest6537 [~Guest65@85.136.168.47.dyn.user.ono.com] has quit [Quit: Client closed]
03:56 -!- juri_ [~juri@79.140.123.169] has quit [Read error: Connection reset by peer]
03:59 -!- juri_ [~juri@79.140.123.169] has joined #hplusroadmap
04:06 -!- juri_ [~juri@79.140.123.169] has quit [Ping timeout: 256 seconds]
04:07 -!- juri_ [~juri@84-19-175-179.pool.ovpn.com] has joined #hplusroadmap
04:41 -!- yashgaroth [~ffffffff@2601:5c4:c780:6aa0:91a6:f133:cf77:cf9d] has joined #hplusroadmap
04:42 -!- balrog [znc@user/balrog] has quit [Read error: Connection reset by peer]
04:43 -!- balrog [znc@user/balrog] has joined #hplusroadmap
05:45 < hprmbridge> Nord> saw your post on e/acc
05:47 < hprmbridge> Nord> Interest in open source hardware, cryptography, nutrition, etc. i work in crypto rn
05:50 < kanzure> cool, what area of cryptography?
05:55 -!- Voyager [~Voyager@075-134-000-004.res.spectrum.com] has joined #hplusroadmap
06:24 < kanzure> welcome Voyager
07:33 < kanzure> jaydugger offers: "analog science fiction" https://archive.org/details/pub_analog-science-fiction-fact?tab=about (1960-2015)
07:37 < Voyager> Which are the most interesting issues in the collection?
07:40 < kanzure> i've asked https://twitter.com/JayDugger/status/1637461717603983363
08:41 -!- cthlolo [~lorogue@77.33.23.154.dhcp.fibianet.dk] has quit [Read error: Connection reset by peer]
08:53 -!- Voyager [~Voyager@075-134-000-004.res.spectrum.com] has quit [Ping timeout: 264 seconds]
09:03 -!- Voyager [~Voyager@075-134-000-004.res.spectrum.com] has joined #hplusroadmap
09:08 < hprmbridge> msnewgooty> For long insertions, scars-based insertion should actually cover a majority of cases I think -- unless you're trying to generate a fusion protein, the scars most likely lie in a non-coding region.  The AttL/R can also act as a linker (although I think that's not necessarily ideal for actual in vivo therapies). But scar vs scarless of course depends on application
09:42 -!- balrog [znc@user/balrog] has quit [Read error: Connection reset by peer]
09:43 -!- balrog [znc@user/balrog] has joined #hplusroadmap
10:06 < hprmbridge> nmz787> Do people still care about "who is satoshi"?
10:08 < Voyager> Doesn't matter much to me. It would be cool to know though.
10:09 < hprmbridge> CheekyMonkey> $27bn is a lot of money and power, but the BTC community usually doesn't want it to come out since can hurt the credibility of BTC
10:10 < Voyager> I guess regardless of who it is, it's an infohazard. 
10:11 < hprmbridge> nmz787> My interest in cryptocurrency has declined logarithmically since it became a popular media topic
10:13 < kanzure> i think what's most interesting about satoshi nakamoto is that it is very strong proof that privacy works and it's still possible to take anonymous action
10:13 < kanzure> it's a valuable message.
10:14 < hprmbridge> kanzure> https://www.nsf.gov/awardsearch/showAward?AWD_ID=2123781&HistoricalAwards=false
10:14 < Voyager> That's never crossed my mind. Interesting way of viewing 
10:15 < hprmbridge> kanzure> https://www.newscientist.com/article/2363095-80000-mouse-brain-cells-used-to-build-a-living-computer/
10:23 -!- estill01 [~estill01@c-73-241-102-105.hsd1.ca.comcast.net] has joined #hplusroadmap
10:43 -!- cthlolo [~lorogue@77.33.23.154.dhcp.fibianet.dk] has joined #hplusroadmap
11:08 -!- Guest60 [~Guest60@2a09:bac5:c533:15e1::22e:3f] has joined #hplusroadmap
11:16 -!- Voyager [~Voyager@075-134-000-004.res.spectrum.com] has quit [Ping timeout: 255 seconds]
11:16 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=bfe3f986 Bryan Bishop: transcript: balaji and the bitsignal >> http://diyhpl.us/diyhpluswiki/transcripts/2023-03-18-balajis-bitsignal/
11:17 < hprmbridge> Perry> I will bet essentially any amount Balaji is wrong and there will not be $1m/bitcoin pricing in 90 days.
11:17 < hprmbridge> Perry> I would have taken his offered bet unconditionally if he had let me.
11:18 < kanzure> oh of course, it's clearly just a marketing budget
11:18 < hprmbridge> Perry> His bet was stupid too. It was a no brainer, free money. Just borrow $25k at 7% to buy a bitcoin, lock in the profit with his offer, get $975k - epsilon in 90 days even if he’s vaguely correct.
11:18 < kanzure> i found him to be rather rambly and uninspired- i was actually hoping for a little bit more brain here
11:18 < hprmbridge> Perry> I like him in general but this is his waterloo.
11:19 < hprmbridge> Perry> I’m disgusted that anyone would help anyone figure out who Satoshi is, or spend so much time trying to fuck over the lives of innocent people who then need bodyguards for the rest of their lives.
11:20 < hprmbridge> Perry> It’s not innocent to say “I think person X is Satoshi”, after that moment that person needs to pay for guards for their kids even if they don’t have any money to do it.
11:21 < kanzure> yes there's a significant personal danger aspect to the satoshi mystery-- even CSW was raided hours after wired magazine ran their article (it's not CSW but he still doesn't deserve to get raided)
11:22 < hprmbridge> CheekyMonkey> if we find out the real Satoshi, he can afford bodyguards, while the fake Satoshis who are speculated against and can't afford bodyguards won't be threatened
11:22 < kanzure> there is actually no indication that the real satoshi would be able to afford bodyguards
11:22 < hprmbridge> Perry> “Satoshi” obviously didn’t want to be found. If someone goes through trouble to get privacy they should be left alone.
11:22 < kanzure> in fact, many of the bitcoin developers were entirely unpaid volunteers who were not in an economic position to invest significant funds into bitcoin
11:23 < hprmbridge> Perry> There’s a reason I deleted all the logs from the cryptography list infrastructure, to prevent any temptation on my own part or anyone else’s to do a dumpster dive through them.
11:23 < kanzure> yeah there's dangling bits of data like subscribers and who subscribed when
11:23 < kanzure> or any records of bounce messages and what was bouncing when
11:23 < kanzure> the p2pfoundation mailing list still has a public member list
11:24 < hprmbridge> Perry> Trying to figure out who “Satoshi” is/was is of no public interest. It’s pure pornography.
11:24 < hprmbridge> CheekyMonkey> if Satoshi does have access to 1.1mm btc as is speculated then that is quite a bit of power
11:25 < kanzure> i wouldn't take that assumption at face value really.
11:27 -!- Voyager [~Voyager@075-134-000-004.res.spectrum.com] has joined #hplusroadmap
11:27 < hprmbridge> Perry> Why is anyone’s identity or life fair game? Whether it’s a pauper or a billionaire?
11:27 < hprmbridge> CheekyMonkey> there's a lot of political influence done through money by billionaires often they try to hide it but often it does eventually come out
11:27 < hprmbridge> Perry> Not your business, not your life.
11:28 < hprmbridge> Perry> You’re hiding behind a pseudonym.
11:28 < hprmbridge> Perry> Want me to dox you?
11:28 < hprmbridge> Perry> If not, then don’t dox others or suggest they should be doxed.
11:28 < kanzure> as a reminder we do have a general review here to not spend too much time on politics- let the billionaires be
11:28 < hprmbridge> Perry> It’s not civilized behavior.
11:28 < kanzure> .. general rule, rather.
11:28 < L29Ah> we need some billionaires spend more monies on h+
11:29 < hprmbridge> CheekyMonkey> Vitalik is a big supporter of longevity
11:29 < kanzure> i think there are billionaires spending money on transhumanist projects but for the fun stuff there isn't a lot of reason to be very public about it
11:30 < L29Ah> i wonder how many clandestine human cloning labs are in africa or pacific these days
11:30 < kanzure> the other issue is that many of the projects are often presented in an unfocused way and don't don't use or fit into the right language to get investment
11:30 -!- Guest60 [~Guest60@2a09:bac5:c533:15e1::22e:3f] has quit [Quit: Client closed]
11:30 < hprmbridge> Perry> They might exist but there aren’t a lot of them I suspect.
11:30 < kanzure> if there really was a cloned child running around or an embryo edited child, there wouldn't be any announcements about who specifically it is- like grimes going for that surrogacy pregnancy with elon on kid number 2 (or 11 depending on how you count)
11:31 < kanzure> it's not like they would be racing to announce it and put a kid in that position for life
11:31 < kanzure> (in fact, putting any children at all into social media is pretty bonkers)
11:33 -!- balrog [znc@user/balrog] has quit [Quit: Bye]
11:35 -!- balrog [znc@user/balrog] has joined #hplusroadmap
11:46  * nsh imagines a redundant array of of independent satoshis
11:46 < nsh> -of
12:04 -!- Voyager [~Voyager@075-134-000-004.res.spectrum.com] has quit [Ping timeout: 248 seconds]
12:25 -!- Voyager [~Voyager@075-134-000-004.res.spectrum.com] has joined #hplusroadmap
12:27 -!- daniel_ [uid593091@id-593091.helmsley.irccloud.com] has quit [Quit: Connection closed for inactivity]
12:39 -!- Voyager [~Voyager@075-134-000-004.res.spectrum.com] has quit [Ping timeout: 250 seconds]
13:04 -!- Voyager [~Voyager@075-134-000-004.res.spectrum.com] has joined #hplusroadmap
13:04 < kanzure> nsh: /me actions don't transmit over discord at the moment.
13:05  * nsh shrugs, smiling
13:22 -!- VividVoyager [~Voyager@075-134-000-004.res.spectrum.com] has joined #hplusroadmap
13:25 -!- Voyager [~Voyager@075-134-000-004.res.spectrum.com] has quit [Ping timeout: 246 seconds]
14:11 < fenn> the analog scifi magazine collection is only 1960-1961
14:30 -!- VividVoyager [~Voyager@075-134-000-004.res.spectrum.com] has quit [Ping timeout: 250 seconds]
14:30 < fenn> "you can react to the image u want to use with the :envelope: emoji to find the seed"
14:30 < fenn> -_-
14:31 < fenn> where is the "murder all humanity now" button please
14:42 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=18ba3c3f Bryan Bishop: transcript: caitlin long v. nic carter >> http://diyhpl.us/diyhpluswiki/transcripts/2023-03-17-on-the-margin-nic-carter-caitlin-long/
14:50 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=1ddb1c2e Bryan Bishop: add tweeter links >> http://diyhpl.us/diyhpluswiki/transcripts/2023-03-18-balajis-bitsignal/
15:10 -!- cthlolo [~lorogue@77.33.23.154.dhcp.fibianet.dk] has quit [Read error: Connection reset by peer]
15:32 -!- balrog [znc@user/balrog] has quit [Quit: Bye]
15:35 -!- balrog [znc@user/balrog] has joined #hplusroadmap
15:49 -!- VividVoyager [~Voyager@075-134-000-004.res.spectrum.com] has joined #hplusroadmap
15:49 -!- VividVoyager [~Voyager@075-134-000-004.res.spectrum.com] has quit [Client Quit]
16:19 -!- estill01 [~estill01@c-73-241-102-105.hsd1.ca.comcast.net] has quit [Quit: Client closed]
18:30 < hprmbridge> an1lam> (post author) personally, the main reason is because we are biology (still)!
18:32 < kanzure> hm which post author?
18:33 < hprmbridge> an1lam> Important to clarify that partly due to my lack of nuance in the post I think it may be getting mis-interpreted. I am very much pro pushing on biological progress. My point is that I am skeptical of the vision of 100s of years from now biological humans being the thing that expands beyond nearby planets because (even with genetic engineering to improve rad resistance) we are so much less robust
18:33 < hprmbridge> an1lam> than non-organic systems. Even if you buy that bacteria are useful for terraforming or whatever, I still think any "precious" cargo (such as humans) is gonna be safer as hardware that you can back-up, easily shield, etc. etc.
18:33 < hprmbridge> an1lam> Stephen
18:33 < kanzure> yeah it's kind of interesting that the zeitgeist right now is kind of "pick carbon or pick silicon but you can't have both" i have no idea where that is coming from
18:34 < kanzure> as a software person i definitely agree biology sucks and it's super messy but computers are also kind of annoying to deal with in their own ways- different problems i guess.
18:34 < hprmbridge> lachlan> it’ll eventually come down to “pick carbon (crystalline) or pick carbon (hydrocarbons)”
18:35 < kanzure> sure but at the very least we should ask people to if not commit to transhumanism then to at least not to slow down organic carbon enhancement research
18:35 < kanzure> (or intentionally block it or whatever)
18:36 < hprmbridge> lachlan> for sure
18:36 < kanzure> i also think a lot of the silicon stuff is going to have huge bottlenecks on the supply and manufacturing side that will limit overall access to compute. biology at least has (almost magical) self-replication capabilities.
18:37 < kanzure> silicon could eventually have self-replication but we currently don't know how to do that. there is no "bitter lesson" of self-replication yet. the "bitter lesson" of self-replication might eventually be "just use biology".
18:41 < hprmbridge> an1lam> I'm starting to suspect that it's partly aesthetics, partly betting on virtual vs. real world (in which case apm folks would also be in latter), partly other stuff I'm not sure about
18:42 < hprmbridge> lachlan> People don’t want future generations to be smarter than them
18:44 < kanzure> we're talking about this one right? https://stephenmalina.com/post/2023-01-11-viriditas-dialogue/
18:45 < kanzure> people reading the backlog, this could bring a tear to my eye if i had any water left in my carbon carcass
18:48 < hprmbridge> lachlan> That’s a very interesting article
18:52 < hprmbridge> lachlan> I admit I have a hard time imagining biology being more than a tool orchestrated by machines in the long-term
18:53 < hprmbridge> an1lam> yeah it feels like in the long term evolution's constraints just don't make sense to stick with
18:54 < hprmbridge> an1lam> so insofar as biology means cells made out of the materials biology currently works with with DNA providing instructions to construct proteins and so on, you eventually will go beyond these things
18:54 < hprmbridge> an1lam> in the short term still very important (imo)
18:57 < hprmbridge> an1lam> at a certain point the binary distinction also probably becomes kind of silly. there's biology as in carbon-based life and then there's the sub-ideas of building out of parts that benefit from or at least handle extremely stochastic environments, multi-level organization, harnessing selection (affinity maturation, directed evolution), which I suspect are all useful engineering ideas for wrangling
18:57 < hprmbridge> an1lam> complex adaptive systems
18:57 < hprmbridge> an1lam> maybe?
18:59 < hprmbridge> lachlan> Humans just aren’t very good at designing things that operate stochasticly and still function
19:03 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has quit [Ping timeout: 246 seconds]
19:12 < hprmbridge> an1lam> True, although presumably in the long term "we" also get better
19:12 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=7d63d4fc Bryan Bishop: transcript: whatbitcoindid with caitlin long >> http://diyhpl.us/diyhpluswiki/transcripts/2023-03-17-whatbitcoindid-caitlin-long/
19:14 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=b940f65e Bryan Bishop: move >> http://diyhpl.us/diyhpluswiki/transcripts/what-bitcoin-did-podcast/2023-03-17-whatbitcoindid-caitlin-long/
19:15 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=1140dc84 Bryan Bishop: tweeter link >> http://diyhpl.us/diyhpluswiki/transcripts/what-bitcoin-did-podcast/2023-03-17-whatbitcoindid-caitlin-long/
19:15 <+gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=2190bbcb Bryan Bishop: transcript: whatbitcoindid with caitlin long >> http://diyhpl.us/diyhpluswiki/transcripts/what-bitcoin-did-podcast/2023-03-17-whatbitcoindid-caitlin-long/
19:15 < kanzure> sorry for the botspam
19:16 -!- yashgaroth [~ffffffff@2601:5c4:c780:6aa0:91a6:f133:cf77:cf9d] has quit [Quit: Leaving]
19:27 < kanzure> after posting on the new e/acc site they had to delete the tweets because hplusroadmap is just too advanced https://www.effectiveacceleration.org/
19:28 < kanzure> https://www.effectiveacceleration.org/posts/KyuwpLJjneLDddpmW/about-the-forum
19:49 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has quit [Ping timeout: 260 seconds]
20:13 -!- balrog [znc@user/balrog] has quit [Read error: Connection reset by peer]
20:15 -!- balrog [znc@user/balrog] has joined #hplusroadmap
20:26 < fenn> so are you just abandoning IRC?
20:27 < hprmbridge> kanzure> what why?
20:27 < fenn> well there's zero mention of it in your announcement post
20:28 < hprmbridge> kanzure> it's in the wiki page? i don't think these people know about IRC.
20:28 < hprmbridge> kanzure> this was the same announcement post from a while back i posted on lesswrong (the one they replied "but won't this cause safety risk" to)
20:29 < fenn> hm i didn't read carefully enough. i thought the link at the bottom was directly to the discord server
20:30 < fenn> sorry, kinda out of it today because of allergies
20:30 < hprmbridge> kanzure> all good.
20:38 < fenn> "effective acceleration" sounds better to my ear
20:38 < hprmbridge> kanzure> don't threaten me with a good time
20:39 < fenn> you don't want to define yourself as a derivative of or reaction to another movement
20:39 < fenn> and accelerationism has some bad connotations already, which supposedly are not intended
20:40 < fenn> also it's hard to proounce
20:40 < muurkha> like Marxism and being associated with Nick Land?
20:40 < fenn> like encouraging genocide and the destruction of civilization
20:41 < fenn> i don't remember exactly, look it up yourself
20:41 < muurkha> connotations are hard to look up
20:41 < fenn> it's what all those clown emojis are referring to
20:54 < hprmbridge> kanzure> yt-whisper https://gist.github.com/jooray/b7266366dbc2f0c162542cf9927768f9
20:55 < hprmbridge> kanzure> https://vid2txt.com/
20:56 < fenn> can whisper dump multiple subtitle languages at once? it seems like recognizing the audio is the hard part
20:58 < hprmbridge> kanzure> i forget his IRC handle (rusnak?) but you could ask this person https://twitter.com/PavolRusnak/status/1637094163916152838
21:00 < fenn> translations with deep understanding of the timing for each word/concept is an underappreciated aspect of whisper
21:01 < fenn> it's something you can't get by dumping pre-recognized sentences into a translator
21:01 < hprmbridge> kanzure> whycome whisper hasn't completely replaced me yet, I want a break
21:01 < fenn> dunno, why?
21:01 < fenn> i thought you liked typing
21:02 < hprmbridge> kanzure> it's slow and doesn't produce as good jargon-informed wording
21:02 < hprmbridge> kanzure> not really
21:02 < hprmbridge> kanzure> got better things to do
21:02 < hprmbridge> kanzure> it's everyone else who has been rude putting out audio without text....
21:04 -!- Mabel is now known as SYCOPIA
21:18 -!- SYCOPIA is now known as ILODEXTR
21:35 < fenn> pretrained alpaca models in the wild: https://github.com/antimatter15/alpaca.cpp#get-started
21:38 < superkuh> I already have an IRC bot running it.
21:39 < fenn> who's that?
21:39 < superkuh> It's not here. I could bring an instance in.
21:39 < fenn> does it respond to PMs?
21:39 < fenn> my only concern with bringing it here is that people tend to spam the shit out the channel when there's a new toy
21:39 < superkuh> To people who are in the channel with it.
21:40 < fenn> perfect, please do
21:40 < muurkha> I'd like to have a chance to play with it myself.
21:40 < muurkha> Is there a channel dedicated to it?
21:41 < superkuh> Mine's real slow because it's CPU and because I just wrap STDOUT/IN to collect the output word fragments as they come out and the delay between tokens is not deterministic no matter how often I add flushing stdout to the cpp.
21:41 < fenn> what sort of prompt do you use?
21:41 < superkuh> Just the default that copies the stanford training set style that in the cpp.
21:42 < superkuh> I'll sanitize the code and put it up in a bit.
21:43 < fenn> i meant for the irc part, not the training prompt
21:44 < superkuh> I'm not sure what you mean. Just the user message.
21:44 < fenn> uh, i guess you mean: "Below is an instruction that describes a task. Write a response that appropriately completes the request."
21:44 < superkuh> Yes, that's built into the interactive mode of alpaca.cpp which I modify and wrap the in/out with perl.
21:46 < superkuh> Okay, it's "kuhbot" in ##OpenAI (don't mind the name, just a place to discuss related things with the most people and play with bots)
21:47 < superkuh> https://github.com/antimatter15/alpaca.cpp/blob/master/chat.cpp#L874
21:47 < superkuh> re: pre-prompt
22:06 < fenn> 6 days ago is eons in AI time
22:06 < fenn> i hadn't updated my alpaca.cpp repo
23:33 -!- ILODEXTR [~Malvolio@idlerpg/player/Malvolio] has quit [Quit: 18FF :all normalized peoples undergone the process of normalization ACGATTTT 2023-03-20 06:33:11:585]
--- Log closed Mon Mar 20 00:00:15 2023