--- Log opened Wed Sep 06 00:00:14 2023
00:00 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has joined #hplusroadmap
01:10 -!- test_ [flooded@gateway/vpn/protonvpn/flood/x-43489060] has joined #hplusroadmap
01:14 -!- flooded [flooded@gateway/vpn/protonvpn/flood/x-43489060] has quit [Ping timeout: 246 seconds]
01:32 -!- Croran [~Croran@user/Croran] has quit [Ping timeout: 255 seconds]
01:38 -!- Croran [~Croran@user/Croran] has joined #hplusroadmap
03:19 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has quit [Ping timeout: 255 seconds]
03:51 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
04:20 -!- flooded [flooded@gateway/vpn/protonvpn/flood/x-43489060] has joined #hplusroadmap
04:23 -!- test_ [flooded@gateway/vpn/protonvpn/flood/x-43489060] has quit [Ping timeout: 258 seconds]
05:19 -!- srk [~sorki@user/srk] has quit [Server closed connection]
05:19 -!- srk [~sorki@user/srk] has joined #hplusroadmap
06:12 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has quit [Ping timeout: 258 seconds]
07:05 -!- stipa [~stipa@user/stipa] has joined #hplusroadmap
07:12 -!- stipa [~stipa@user/stipa] has quit [Ping timeout: 246 seconds]
07:20 -!- stipa [~stipa@user/stipa] has joined #hplusroadmap
07:29 -!- test_ [flooded@gateway/vpn/protonvpn/flood/x-43489060] has joined #hplusroadmap
07:32 -!- flooded [flooded@gateway/vpn/protonvpn/flood/x-43489060] has quit [Ping timeout: 246 seconds]
07:44 < docl> https://www.frontiersin.org/articles/10.3389/fceng.2020.612594/full
08:19 -!- stipa [~stipa@user/stipa] has quit [Quit: WeeChat 3.0]
08:45 < kanzure> .title
08:45 < EmmyNoether> Frontiers | Minicircle Biopharmaceuticals–An Overview of Purification Strategies
08:49 -!- cthlolo [~lorogue@p549d1eb6.dip0.t-ipconnect.de] has joined #hplusroadmap
08:54 -!- cthlolo [~lorogue@p549d1eb6.dip0.t-ipconnect.de] has quit [Remote host closed the connection]
08:55 -!- cthlolo [~lorogue@p549d1eb6.dip0.t-ipconnect.de] has joined #hplusroadmap
09:03 -!- cthlolo [~lorogue@p549d1eb6.dip0.t-ipconnect.de] has quit [Remote host closed the connection]
09:03 -!- cthlolo [~lorogue@p549d1eb6.dip0.t-ipconnect.de] has joined #hplusroadmap
09:05 -!- cthlolo_sec [~lorogue@77.241.128.194.mobile.3.dk] has joined #hplusroadmap
09:08 -!- cthlolo [~lorogue@p549d1eb6.dip0.t-ipconnect.de] has quit [Ping timeout: 245 seconds]
09:16 -!- cthlolo_sec [~lorogue@77.241.128.194.mobile.3.dk] has quit [Quit: Leaving]
09:42 < hprmbridge> kanzure> https://spec.tech/library/macromolecular-additive-manufacturing-roadmap
10:01 < hprmbridge> kanzure> what are "the nexus bots" ? just saw a reference to this.
10:04 < docl> .title https://nexus.bot/
10:04 < EmmyNoether> Nexus | Discord Bots as a Service
10:05 < hprmbridge> kanzure> https://www.reddit.com/r/scihub/comments/13cms8m/how_to_use_nexus_bots_or_stc_to_download_the/
10:09 < hprmbridge> kanzure> https://www.reddit.com/r/scihub/comments/12detqs/standard_template_construct_store_and_search_the/
10:15 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has joined #hplusroadmap
11:08 < hprmbridge> kanzure> from https://twitter.com/utxoclub/status/1699479343800443198 https://cdn.discordapp.com/attachments/1064664282450628710/1149043379544527020/F5W8jPEWYAAiqIk.png
11:57 < hprmbridge> Eli> Cool. Have you tested this yet?
11:59 < kanzure> no. you?
12:00 < kanzure> "Complete human day 14 post-implantation embryo models from naïve ES cells" https://www.nature.com/articles/s41586-023-06604-5
12:00 < hprmbridge> alonzoc> Huh I was tinkering with something along these lines myself with my private library
12:02 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has left #hplusroadmap []
12:12 < hprmbridge> Eli> I’ve been using Anna’s archive
12:12 < hprmbridge> Eli> But, a backup would be super helpful
12:12 < hprmbridge> Eli> https://annas-archive.org/
12:15 -!- test_ is now known as _flood
14:38 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
14:41 < docl> .title https://www.youtube.com/watch?v=vNSd_FcdDII
14:41 < EmmyNoether> Andrew Turberfield, Chris Wintersinger, Erik Poppleton | Molecular Additive Manufacturing - YouTube
14:52 -!- NewtonTrendy [~ubuntu@user/bopqod] has changed host
14:56 < hprmbridge> alonzoc> Anyone got resources on optimal training plans for fully body workouts? I assume someone somewhere has put research into optimised regimens for functional strength etc?
14:57 < hprmbridge> alonzoc> *full body
15:02 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has quit [Ping timeout: 246 seconds]
15:18 < TMA> alonzoc: if someone did, it did not leave a noticeable dent in the memesphere which remains saturated with snake oil purveyors
15:18 < hprmbridge> alonzoc> That's my problem
15:20 < TMA> alonzoc: my guess is that it is hard, because of the variability of people is still poorly understood and a regimen that works optimally for certain subgroup will have disastrous effect on other subgroup
15:21 < hprmbridge> alonzoc> I guess, however you think that you could have your training protocol be based on a persons current capabilities and state, then instead of a "week 2 you do this" you have "given these performance metrics as this week's baseline you do xyz"
15:24 < hprmbridge> alonzoc> Hmm I wonder if you recorded bulk data of lot's of people's fitness training and capabilities you could learn a causal model that predicts that data and then run MCTS on that model to generate a workout plans on the fly given someone's inputted utility function and current data.
15:24 < hprmbridge> alonzoc> Hmm someone ought to make some fitness startup based on that idea
15:25 < hprmbridge> alonzoc> A chain of "data optimised" gyms or something
15:25 < hprmbridge> alonzoc> Lol SV VCs would prolly bite
15:29 < TMA> Is it even doable? some people can get a significant proportion of the excercise benefits just by imagining to excercise
15:33 < TMA> I mean: is the variability of gym-goers low enough in order to gather enough data to generalize from while remaining economically viable?
15:34 < hprmbridge> alonzoc> Well there's natural variation amongst people be it proportion of type 1 v type 2 muscle fibres or whatever.
15:34 < hprmbridge> alonzoc> However all of those are hidden variables, so if you record someone's past training history you can try an infer hidden variables in the model. So if you're model class is sufficiently expressive it could implicitly pick up on each individuals specific variations
15:34 < hprmbridge> alonzoc> It basically becomes a bayesian filtering task
15:35 < TMA> the amount of data needed goes up with each hidden variable, otherwise you are learning from a signal that is drowned in noise
15:37 < hprmbridge> alonzoc> The issue is data volume which is why you'd need to open a chain of gyms etc and it'd have to be a startup with VC funding. However the age of careless money slinging is apparently closing
15:38 < TMA> it was never quite open where I am located
15:38 < hprmbridge> alonzoc> But even a low order fit at first would give a decent fitness plan that'd work for most
15:39 < hprmbridge> alonzoc> Well I'm from the UK so same boat
15:39 < hprmbridge> alonzoc> Would've loved to walk into some VC and pitch "tinder for dogs" or whatever
15:40 < TMA> a decent fitness plan is anything in the "excercise, drink fluids, eat balanced diet in moderation, rest well" ballpark
15:40 < TMA> not optimal nor tailored though
15:42 < TMA> alonzoc: that would be funny, especially after pivoting to "iFlorist advice app"
15:46 < hprmbridge> alonzoc> Tbh I don't think I'd enjoy working on a useless app or some such. 99% of bigtech startups are a simple formula of "isolate some need or want and the parties involved. build a platform to connect parties in a network, and maximise the ease (make frictionless) and thus "volume" of those transactions. Scale until you capture the market"
15:46 < hprmbridge> alonzoc> E.g. Youtube. Want: to watch content. Parties: consumers and creators. Transaction: watching a piece of content
15:46 < hprmbridge> alonzoc> Tinder. Want: to hookup. Parties: People in general. Transaction: a date
15:46 < hprmbridge> alonzoc> Uber. Want: transport. Parties: people and drivers. Transaction: a ride
15:46 < hprmbridge> alonzoc>
15:46 < hprmbridge> alonzoc> It all comes down to isolating a want and it's associating transaction, and then transplanting the associated "network" to a centralised platform
15:47 < hprmbridge> alonzoc> They're all just a fancy UI and a matching engine basically. Pretty much every one. Amazon, Google search, etc.
15:48 < hprmbridge> alonzoc> /end rant about how all bigtech companies use the exact same formula
15:51 < hprmbridge> alonzoc> Netflix even though it's basically a standard cable company. And their lack of scalability and thus growth is the fact half their network are entrenched media blobs and their transactions involve expensive content creation they need to pay for. While YouTube has a effectively bipartite user pool but they don't need to pay either to come to them
15:51 < hprmbridge> alonzoc> Well aside from ads but you get the point
15:54 < docl> this doc is amazing https://spec.tech/library/macromolecular-additive-manufacturing-roadmap
15:54 < TMA> in a sense the business model is flawed. because the "maximise the ease (make frictionless)" is hampered by the ads, which add friction
15:56 < hprmbridge> alonzoc> Well once you've captured the associated network for the want you're targeting you can put in ads because you've got a monopoly. The whole VC make losses for a while game is because you pour all the money into capturing a huge network you can then exploit because the cost of leaving the platform is a massive loss in convenience. Made worse because everyone just uses the platform now.
15:57 < hprmbridge> alonzoc> It's why you get "amazing" new apps that start out good and then go to shit once they've matured and captured an audience. *Cough* YouTube *cough* Spotify *cough* etc
15:59 < docl> diseconomy of scale?
15:59 < hprmbridge> alonzoc> The thing is a lot of them can just be replaced by a federated alternative there's no actual need for the matching engine and easy UI to be centralised into a company. Once the products matured very little needs to be done
16:00 < hprmbridge> alonzoc> Yeah basically
16:00 < docl> I think part of the issue is they get more money committed that can't be disvested without a tax hit. it then becomes acceptable to perform at around the level of competitors
16:01 < hprmbridge> alonzoc> Like if everything was federated the only real change would be ad companies would need to change how they function
16:02 < hprmbridge> alonzoc> But like modern YouTube is full with channels taking sponsorships so platform served ads might not happen at all
16:03 < hprmbridge> alonzoc> Anyways I sound like a grumpy old man, that doc you shared looks interesting
16:04 < hprmbridge> alonzoc> I'm skeptical of biological bootstrapping of dry nanotech but giant proteins have tons of applications before you get anywhere near that
16:08 < hprmbridge> Eli> This is what personal trainers are for. The first thing they ask you is essentially what your objective function is. Bodybuilders have totally different workouts than powerlifters, for example. I'm assuming your goal is lifespan extension, low risk of injury, and maybe some more lean mass? For that, they will devise a custom fitness and nutrition plan for you. This will be customized around your
16:08 < hprmbridge> Eli> body type, any injuries or health concerns you have, and your personal schedule. Then the trainer will ensure you have good form and use jedi mind tricks to push you farther in your workouts.
16:10 < hprmbridge> Eli> It's probably difficult to scale this at the moment. Maybe in the future with VR headsets and robots that can follow you in the gym
16:24 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has quit [Ping timeout: 246 seconds]
16:57 < hprmbridge> alonzoc> Yeah but I'd rather not get a personal trainer. And I'm more interested on proper non snake oil literature. Issue is it's hard to search, if you do Google scholar you get a lot about for patients recovering, the obese or elderly etc but little about standard fitness routines. And if you search the web proper you get people selling their 5 week plan with appropriately garish branding
17:00 < hprmbridge> alonzoc> Might try find some literature on military fitness training, they've not got the same goals as me but they've got the right incentives in designing the PT routines. (And I don't mean the sluge books pumped out by people who left the military)
17:06 < hprmbridge> alonzoc> Tbh I'd like to just solve for constriants on my nutrition and then generate a meal plan via constrained optimisation. Nutritional constraints with optimisation objective being variance and high subjective rating of each dish
17:08 < hprmbridge> alonzoc> I know a guy who basically wrote a giant prolog blob to plan out all his daily tasks and what not. I wouldn't go that far but you could quite easily generate custom meal plans given a set of recipes and their nutritional data
17:24 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
17:46 < fenn> "bacterial ghosts are empty bacterial cell envelopes that are obtained by controlled expression of the cloned lysis gene E of bacteriophage ϕX174 which codes for a membrane protein. In Gram-negative bacteria, this protein oligomerizes into a transmembrane channel structure that lyses the cell and releases its cytoplasmatic content, without compromising the structural integrity of the cell
17:47 < fenn> envelope"
17:47 < fenn> neato
17:49 < fenn> they're used as adjuvants and to "deliver DNA-encoded antigens to dendritic cells" (why "dendritic cells" specifically?)
17:50 < fenn> i guess this is the least cytotoxic way of presenting an antigen to the immune system
17:59 < fenn> heh "standard template construct" is the leftover terran AI technology from the golden age of warhammer 40k
18:01 < fenn> something like a heavily curated and standardized SKDB
18:05 < fenn> https://philipsibbering.com/warhammer/40k/imperial-tech/
18:07 < fenn> "if we, on our world obtain an STC and could understand it, we do not have the machines to make the higher technologies, and we do not have the machines that could make the machine to make the higher technologies. Obviously building machines, to build other more advanced machines, to build even more advanced machines is expensive and not much in the way of capital return, so we would have to walk
18:07 < fenn> the path of the STC, going through each generation and working our way up. However, each step would be mapped out, progress would be accelerated, and it would all work perfectly and be the best that we could imagine, and any improvement we might make would reveal themselves to be false as we progressed along the path"
18:09 < hprmbridge> Eli> Because humans are all very different and have different goals there isn't really a standard fitness routine. You basically choose what you are aiming for and your routine is oriented for that. My guess is that you are looking for more of a functional training workout, if you are a standard white collar worker who is looking at lifting weights. A trainer would be helpful to ensure you aren't
18:09 < hprmbridge> Eli> overcompensating or having bad form. This is the scientific way to workout. If you are lifting incorrectly, the physics of it are not efficient.
18:09 < hprmbridge> Eli>
18:09 < hprmbridge> Eli> For diet, you could use the free version of https://cronometer.com/ and ensure that you are getting a lot of protein. A lot of nutritional epidemiology is just religious cults fighting each other. It's not really a science.
18:09 < fenn> how about recording yourself lifting with a phone duck taped to the wall, and having AI figure it out
18:10 < fenn> surely there's no magic here
18:10 < hprmbridge> Eli> yeah, or in the future just have a tesla bot follow you around and give recommendations. That's probably coming.
18:11 < fenn> lots of people exercise, there must be a large number of videos of people exercising with good and bad form. turning video into pose estimation is a solved problem
18:11 < hprmbridge> Eli> There is also a big mental aspect to the personal trainer side of things. So, it would have to be seen if the bot could properly motivate a human to max out.
18:11 < fenn> flex those meat appendages, organic
18:31 < fenn> did crossref (library index) successfully switch to being not-evil? https://www.crossref.org/blog/2022-public-data-file-of-more-than-134-million-metadata-records-now-available/
18:31 < fenn> (200GB torrent)
18:32 < fenn> maybe i'm confusing it with OCLC
18:41 < fenn> it looks like professional librarians can't keep their blogs online for more than 10 years either
18:56 < fenn> superkuh: thoughts? https://izihawa.github.io/summa/blog/mitigating-internet-censorship-and-privacy-issues/
18:57 < fenn> i can't seem to actually, you know, run the damn thing
19:07 -!- ANACHRON [~Malvolio@idlerpg/player/Malvolio] has quit [Read error: No route to host]
19:10 -!- Malvolio [~Malvolio@idlerpg/player/Malvolio] has joined #hplusroadmap
19:17 -!- Malvolio [~Malvolio@idlerpg/player/Malvolio] has quit [Ping timeout: 258 seconds]
19:24 < superkuh> I don't like running applications in browsers. I feel like ipfs suffers from the same problem as any privacy overlay: the gateways end up censoring. Not a problem for us, I suppose, but for getting widespread use.
19:24 < superkuh> Cloudflare has kind of embraced and extended IPFS to a slow neutering.
19:38 < L29Ah> ipfs is just rebranded NIH bittorrent
19:39 < L29Ah> i'm surprised cloudflare touched it at all
19:47 < fenn> bittorrent has issues with mutable data, like a collection you periodically add to
19:47 < fenn> for dumb reasons, bittorrent hashes on sorta arbitrary chunk boundaries, rather than file hashes
19:49 < fenn> it would be nice if torrent clients were also IPFS nodes
19:49 < fenn> there's no good reason to divide the dataset by the protocol used to share it
19:50 < fenn> i never bothered to learn how to use IPFS because it sounded complicated
19:51 < fenn> hopefully they have fixed that in the last 10 years
19:55 < L29Ah> ipfs still fetches executables and runs them to migrate your local ipfs database to a newer format
19:55 < L29Ah> well, the reference implementation does, they renamed it to "kubo" iirc
20:02 < fenn> gross
20:05 -!- Mabel [~Malvolio@idlerpg/player/Malvolio] has joined #hplusroadmap
20:06 -!- flooded [flooded@gateway/vpn/protonvpn/flood/x-43489060] has joined #hplusroadmap
20:09 -!- _flood [flooded@gateway/vpn/protonvpn/flood/x-43489060] has quit [Ping timeout: 258 seconds]
22:00 -!- stipa [~stipa@user/stipa] has joined #hplusroadmap
22:09 -!- Mabel [~Malvolio@idlerpg/player/Malvolio] has quit [Ping timeout: 258 seconds]
23:16 -!- Malvolio [~Malvolio@idlerpg/player/Malvolio] has joined #hplusroadmap
23:17 -!- Malvolio [~Malvolio@idlerpg/player/Malvolio] has quit [Remote host closed the connection]
--- Log closed Thu Sep 07 00:00:15 2023