2w ago

also expensive

didn't find the post link again, so here is the account https://infosec.exchange/@Em0nM4stodon

You're viewing a single thread.

122 comments

A friend of mine asked me today if there were tech companies I was excited about. The context was more "companies that will grow" not "companies that are doing something cool". But, I was stumped because I had trouble thinking of anything in either category.
Looking at the MANA MANA (do dooo do do do) group:
Microsoft: Always shitty assholes, but their stock price will probably keep going up until the AI bubble pops
Apple: Nothing innovative since the iPhone, but their stock will probably keep doing well because of their duopoly status and the 30% rake on the App Store
Nvidia: I used to like their video cards, but they haven't done anything innovative for gamers since ray tracing, and even that is barely used. When the AI bubble pops they're going to crash hard
Amazon: Assholes who screw over anybody who sells things through them, abuses their employees, and the last "innovation" they had was their patent on one-click ordering. Since AWS is most of their revenue, when the AI bubble pops their revenue will crater.
Meta: Renamed from Facebook because their thundercunt of a CEO thought the future was "the metaverse", an obviously bad idea from the start. The company only continues to be relevant because network effects cause FOMO and they have an advertising duopoly with GOOG, heavily betting on AI now, and will crash when it crashes.
Alphabet: Their flagship service is terrible now, but they don't care because they have such an overwhelming monopoly on search. More importantly, they're part of a massive ad duopoly with Meta, so as long as they can keep you coming back, they'll keep making money. I can't remember them having any innovative ideas since PageRank back when they were founded. They're also all in on AI and will crash when it crashes.
Netflix: It used to be that you only needed 1 streaming service, and it was Netflix. Now the Netflix catalogue is mediocre, and they're getting rid of things that actually made people like them, like allowing a family to share a password, and a truly ad-free experience. I don't see Netflix growing much in the future, and with how bad streaming is becoming, I expect more people to pirate instead.
Adobe: You used to be able to own photoshop, and it was a good product. Now you have to rent it, and they're not even fair and honest about how the rental works. Acrobat Reader used to be a useful free utility. Now they keep enshittifying it. Will they keep making money, probably. Probably won't crash too hard in the future either, although they're a tech stock so when the AI crash happens they'll take some damage too.
It genuinely used to feel like many of the big tech companies were trying to solve problems for end users. Sure, they wanted to make money at the same time, but they actually did provide good services. Google search used to be unbelievably good. It would find the one page on the whole Internet that was the best one for your search. If what you wanted wasn't in the first 10 links, it probably didn't exist on the Internet.. Even when it had ads, the ads were small, clearly marked, and didn't crowd out the actual search results. Netflix had a great catalogue and a great UI and zero ads so it was worth paying a bit and not pirating. Paying a Netflix subscription used to feel like sending a message to the Old Media companies that they were dinosaurs who were on their way out. Apple's iPod and iPhone were really game changers. These days it doesn't seem like any of them really want to make your life better. Instead they want to act as a rent-seeking middleman between you and whatever you want.
After thinking about it for a few minutes, the only for-profit company I could think of that was doing innovative things that made life better for its end-users was Framework. I love that they're trying to make modular laptop, and now an innovative desktop. But, there have got to be others out there I'm forgetting, I hope!
- I'm excited for peer to peer technology, because it brings us closer to what the internet was originally supposed to be like.
  I've recommended Keet (chat app) a bunch of times on lemmy earlier, which works really well and that is cool, but that is just a showcase of what's possible with p2p.
  Streaming media, sharing files, communication, browsing wikipedia, etc etc - this can be done without spying middlemen or data centres in between. Some cool demos here 09:45 https://youtube.com/watch?v=BTCsSwCpGP8&t=776
  
  One thing that seemed interesting in that vein is the Dat software / protocol, and the Beaker web browser.
  The aim was basically to create a distributed, peer-to-peer web. When I saw a presentation on it, I thought "hmm, if this works it will be really cool, but I don't think this is going to take off". It seems I was right because the Beaker browser is now gone, and Dat doesn't seem to be getting updates anymore.
  But, I still think there's hope for a distributed web. It just needs something like a killer app.
  
  Same!! P2p and self hosting is getting better and better!
  I've been searching for an alternative chat platform for a while now and I'm yet to find anything I think I can use with friends and Grandma alike, ya know? 😅 so hearing about this p2p keet app got me really excited!
  Sadly, after a bit of reading and such, I'm not so sure... 😕
  play-store or github seem to be the only install methods
  the github is release-delivery only; source code doesn't appear to be public?
  Keet uses Holepunch's (the company behind keet) "HyperDHT", a distributed hash table, to connect peers. So it seems that, while the comms themselves might be p2p, the app still relies on some server(s) to facilitate their initial connection.
  good news (kinda) though! You can self host a 'p2p server'! But the phrasing on that doc page reinforces that the network itself isn't fully p2p= "Creates a new server for accepting incoming encrypted P2P connections"
  Installed it anyway just to see. Immediately prompted to enable Google's push notifications via MicroG 😭
  the splash page of the app proudly announces "no servers!" - documentation says otherwise 😕
  creation of a username first checks whether the username is available.... Where is that being checked? No servers, right? 🤔
  I want this to be cool, but no source code and foggy talk about servers has my sus-dar goin off a little 🤔 if anyone knows more I'd love to be persuaded!! The app itself is definitely very beautiful and responsive 🙂
  
  You raise a lot of points here, I recommend you join the community room in the app, you'll get every detail from the developers there.
  they haven't opensourced it yet, but they say they will do so, and they have done so with all the components that keet is built on top of. So given that track record, I think it's just a matter of when.
  I asked a developer about the dht, in this context a "server" is just a dht node that you can connect to with its public key (but agree it's confusing they use the same word). the wording might be confusing, but its definitively not what anyone understands as a server in a centralized network https://en.m.wikipedia.org/wiki/Distributed_hash_table
  as i've understood, all push notifications on android has to pass through googles servers (but they are encrypted)
  and they don't need a server to check for duplicates in usernames
  so I recommend you continue to explore and ask around in the chat rooms, figure out if this is for you!
  
  Cheers for the response! Extremely excited to hear you've heard about open sourcing from the devs, I'm gonna keep my eye out for sure!! Excited to read about dht!
  Thanks again for sharing the app! 🙂
  
  yea🤘 the tech is really fascinating. Like yea, the p2p-approach introduces some new challenges, but it solves so many existing ones:
  For example costs. The more popular an app gets, the more traffic it gets, the more it costs to run it. I've heard telegram spends hundreds of millions of dollars on servers, with hundreds of developers.
  P2P is the complete opposite. Keet is made by a small team, and the more people use it, the better it runs (because more peers can relay data). It can scale with no such restrictions.
  someone should do the math of what would be the environmental impact if all communication went p2p instead of datacentres.
  
  Yeah I have been trying to read a bit more about DHT (good lord these are complicated, one video attempted to explain hypercubes??! 😵‍💫). It seems one of the bigger use cases is in torrenting! Which is fascinating, both from technical and security perspectives.
  From what I've learned, it's clear DHT is extremely scalable and resilient, which kicks ass! If it also brings inherent security benefits, I'd say this is a clear choice for a new messaging platform!! 😃 I'll have to learn a bit more first though to be sure.
  What I can say is the app itself is GORGEOUS, and very responsive! The devs are also quite active in the community chat room, and seems to listen to (and have full intents to act on) user feedback, which is amazing!
  
  Oh absolutely. I like the qualitative way they interact with their users. Instead of lots of static pages with lists of issues to vote on, roadmaps, FAQs and that kind of thing, feedback and updates all happen in the chats, interacting with the actual developers. When I make requests or report bugs, ppl chime in and those things actually get addressed, and sometimes fixed really fast. Feels like a digital village!
- Saving this post because it sums up exactly how I feel
- they haven’t done anything innovative for gamers since ray tracing
  Unreal Engine's Lumen (and equivalents in other engines like Cryengine) made 'full' RTX obsolete. I can look at random lighting in Satisfactory that looks like modded Cyberpunk 2077 now. Even full path tracing in 2077 (which runs at a slideshow for me, but I tested experimentally) is just... not really worth it, with everything the performance budget GI saves could be used for instead.
  So there's that, and that's a pretty cool software innovation.
  Honestly that's where the neat stuff is now; outside the huge companies. Especially in software.
  
  I hadn't heard about Lumen, but I'll look it up, thanks!
  
  Look up Crytek's SVOGI! DF did a fantastic, and brief, showcase: https://youtu.be/w31Ct25gJlI?t=446
  Cryengine is back.
  
  Thanks for the recommendation, I added it to my queue.
- I can't remember [Alphabet] having any innovative ideas since PageRank back when they were founded.
  Oh come on, they made Google Wave, that was pretty neat! And... Um... That's it I guess?
  
  There are so many good ideas in the Google back catalog, it feels criminal not to just link to the graveyard.
  From AngularJS to Google Cardboard to Project Ara, really can't express how many genuinely cool ideas they floated and then smothered over the last 20 years.
  
  I never used Google wave, but it really didn't seem all that useful to me. But maybe it was innovative? I dunno.
- Is (non-neuralink) deep brain simulation interesting because I know some doctors and they probably know some companies. Never asked to get dad's cyborg parts back when he died for some reason.
  
  It would be interesting if it actually works. It's really promising, but it still seems like it's something that will be cool when it happens at some point in the future, rather than something that is happening now.
  
  i mean it worked in my dad. he was part of a trial to install DBS on moderate parkinson's patients rather than waiting until the patients had severe parkinson's. Short story, gave him ten extra years he could work. A bit longer and more details, he was able to manage nearly all of his dyskinesia through the implant rather than via medication (some kind of levi/carbidopa). It was a really neat device, the MDs who put it in were the best at what they do (and, as a professional patient I've gotten good at evaluating that) and provided us with all the support we needed up until dad died. So our experience was nothing but positive. I think the charger is in the garage and I can dig it up tomorrow to find out what company built his computer if you want.
- I really had to dig through my cynicism to the buried tech optimist in me. But here is some tech I think is really cool.
  Framework: obviously
  Cloudflare: the features like cloudflare workers and anti ai stuff is pretty cool. Ddos mitigation they do is impressive.
  Nvidia: their GPU tech is insane. They are going full stack with their own networking and GPU on the motherboard like a CPU. Their dlss as much as I hate it is very impressive.
  Perplexity browser is interesting to me I can't wait to see what it turns out to be. The idea of having a new way to browse the web is cool.
  Ar glasses are getting really good x real air 2s i want a pair so bad.
  Self driving cars: waymo and Tesla self driving is incredible.
  Boston dynamics robots are sick. Warehouse logistic robots are sick I really like what Amazon is doing on that front.
  Bluesky is cool tech as much as I hate that it copied the fediverse and usurped it with vc funding
  
  Yeah, Cloudflare is doing some interesting things. But, for the most part those aren't consumer-focused services.
  Perplexity is one of the worst of the AI offenders. Their crawlers don't respect "robots.txt" or other things that say that LLM crawling isn't allowed.
  For self-driving cars, I'll give Google credit there. Their Waymo division is really making progress in self-driving cars. They didn't come up with the concept, but they're pushing the boundaries of what's possible.
  OTOH, Tesla's self-driving is a joke. In fact, by calling their bullshit "full self-driving", they've forced the legitimate self driving car companies to use a different term. Tesla's self driving is so bad that it's hurting the rest of the industry and setting back the possibility of actual self-driving vehicles by years.
  Boston Dynamics humanoid robots are cool, I'm not as impressed with their robodog though. But, from what I saw from Beijing last week, they're already way ahead of Boston Dynamics. Even when some of their bots were failing, the kinds of movement they were making before they failed seemed more advanced (and natural) than the Boston Dynamics bots.
  I disagree with some of your choices, but you've got some good ideas too.
  
  Perplexitys methods are bad but the concept of a natural language web browser is a completely unique thing and thats interesting to me. Its like when google enabled people to search "how to change a tire" instead of before where people searched with keywords.
  Idk how you can say self driving tech is not cool. Even the worst self driving which is Tesla's is still very impressive. Telsa can drive autonomously through the street using only cameras. Is it perfect no but its still very good and its only a joke because of how good the competition is. That gives me a lot of hope for self driving cars.
  As for that Chinese robot dog yeah I see the video that got put out, I dont think it blows away Boston dynamics version in anything but manufacturing costs. All the falling and all terrain movement the Chinese dog can do has been done by Boston dynamics for almost 5 years at this point.
  
  Self driving tech is cool. Tesla's take on Self Driving is not cool because it's not effective.
  Telsa can drive autonomously through the street using only cameras.
  Sorta... a bit... in a way that will lead to an accident sooner or later. If they put LIDAR on their cars it would be far more effective, but Musk wants to be different. He insists on only using cameras even though you can't safely do self-driving with only cameras. Typical Musk, cutting corners and lying.
  
  Yeah its dumb to not use lidar and elon musk sucks but the cars can self drive without a doubt. That self driving is impressive even if waymo is already doing taxi services with near perfect driving for the past few years.
  
  the cars can self drive without a doubt
  So can my sister's car for a few seconds if you put the cruise control on. But it can't self-drive safely, and neither can Teslas. But, my sister's car doesn't advertise the ability to self-drive. But, Musk pretends that Telsas can, which is extremely dangerous. He's killing people by muddying the waters and pretending his cars can self-drive safely.
  
  Ok idc about defending tesla anymore. Self driving is cool tech and we can pretend Tesla's cant self drive.
  
  Sounds good.
- Oh, and my controversial take:
  Framework
  I feel like Framework started something awesome and... is stalling?
  The silicon they use is getting a little long in the tooth, and so is the engineering of the cooling, the screen quality... I get it, they're a scrappy startup, but it almost feels like they're stuck.
  Meanwhile the Framework Desktop has awesome hardware, but is largely non modular by necessity and... not available in a laptop? And not very expandable as a desktop, not even with a dGPU slot. And expensive.
  
  I haven't been following them that closely. I hope they come out with new stuff soon though, because I really want them to succeed. Mostly, I want this concept to succeed though. So, if they stumble, I hope someone else picks up the baton.
- Apple died with Steve Jobs. They went from being a company whose success was based on making things that people wanted to becoming a company that only cares about “maximizing value for shareholders.” Having customers is now just an inconvenience.
  Late stage Capitalism in action.
  
  Apple died with Steve Jobs.
  Steve Jobs was a psychopath. He had maybe two good ideas (both of which Microsoft did first) and a ruthless drive to hustle those ideas into the public consciousness. But Apple was, at its heart, an advertising company that made some useful technology. It was so much of an advertising company that Jobs ended up dying from his own kool-aid, convinced he could outsmart the nation's leading oncologists when he was diagnosed with an easily treatable form of cancer.
  I only hope Musk and Zuck suffer the same fate.
  
  It also feels like they’re trying to be like Steve but without any creativity.
  I don’t think he’d ever have thought VR was a big deal, for example.
  
  We know that Steve considered AR to be the next major leap in personal technology, but that vision has almost nothing in common with Vision Pro. AR should be able to change and enhance your environment, not separate you from it. Even if you consider Vision Pro a demonstration of the concept owing to the fact that the technology to make it properly doesn’t exist, it fails at even that. It fails at every meaningful use case of AR. AR is not “phone apps floating in space”, it’s about recognizing and augmenting the world around you, the objects around you, and the actual physical space that we inhabit. I’ve given up on ever seeing proper AR in my lifetime, even as a bulky, ugly, expensive proof of concept.
  
  Yeah, and I think Steve would have thrown any VR headset prototype across the room and fired everyone involved.
  I agree with you on AR. The power requirements of computing have gone way down, but making a display bright enough to see in daylight needs a lot of power, and making them light enough to last all day would require massive improvements in battery tech.

122 comments