I just learned one thing about #Peertube that can make it quite attractive to content creators. Peertube channels have #podcast feeds, and most podcast apps can play video podcasts these days. You can submit the feed to Spotify or Apple Podcasts, making it available to millions of users. AFAIK, you can't do that with videos uploaded to #YouTube channels. I haven't tried if it really works. But if it does, it's a big deal. This is how @mestskacyklistika looks in the #Podverse app.
Calling all Python expert friends and Masto peoples who also know something about NVDA add-ons. Right now, the BeSTSpeak TTS thread can still on occasion crash with certain speech interrupts, particularly if larger chunks of text are fed at once to the TTS. I honestly don't have the knowledge to diagnose all possible combinations of thread management, and all of this is reminding me too much of my C++ college days of race condition handling and such. GPT Pro mode gave me like 7 ways to diagnose it, I just don't have the energy, nor the expertise myself in Python, so if someone picks it up, be my guest. Possible solutions detailed in eurpod.com/crashes.txt and the add-on at eurpod.com/BestSpeak.nvda-addo…
I think it is a pretty good addon. I have noticed if I do things fast and keep interupting it, it will either spell everything, or it makes a really bad static sound. the control key stops that one. for the spelling of everything. I just shut down NVDA, and restart it, and that fixes that one. still very nice to hear this from the classic days.
yeah, some of that may get ironed out in the next few days :) I probably won't have the energy to pick it up again this hardcore tomorrow, like 8 or 9 hours of just coding with GPT straight and barely looking up to go to the bathroom twice. Aha. This is the mess any kind of coding projects get me into. Seeing others from the community step in though gave me a bit of hope that I'm not alone on it and let's face it, there exist far better Python experts than myself out there who could own me and my GPT crafted knowledge any day haha.
TIL, Pemerintah Indonesia selepas era orde baru tetap melarang penyebaran paham komunisme karena demokrasi negeri ini berkembang bersama dengan sistem oligarki dan kapitalisme. Karena itulah, ada upaya oleh orang-orang berpaham kapitalisme agar ajaran paham komunisme tidak dapat berkembang di negeri ini.
Walaupun RKUHP membatasi kriminalisasi apabila dilakukan untuk kepentingan ilmu pengetahuan, tidak berarti setiap orang dengan leluasa mempelajarinya, apalagi mengembangkannya.
Shoutout to @masonasons for helping to add headSize and excitation options to the BeSTSpeech add-on :) Now you can make that Keynote Gold sound whispery or like a chipmunk, if you wanted to. Also had help with improving the capitalization pitch shift function so it should not get stuck and keep spelling out everything in certain situations. Love to see bugs getting squashed. (eurpod.com/BestSpeak.nvda-addo… for those who do not have the link)
@austinnix This is the coolest thing ever. I’ve been looking for Keynote for decades now. The last time I used it, was when I owned a braille note classic.
@BTyson ooh nice, thanks for adding it to the ring, it does show here :) I updated the copy in the main add-on to have this one as this is really useful to have. I tried to make a modified version that didn't do setting explicit argtypes for the DLL functions, since that's how we initially had it when I began using the code @TheQuinbox gave using BGQueue. Ctypes guesses argument types and passes integers as needed, which is really not great practice, but it could be that in the In the new code, us setting bstShutup.argtypes = [ctypes.c_long] causes a bad mismatch. According to the typedef I found in the repo, bstShutup expects a long* (a pointer), not just a long. Nevertheless that did not solve the weird crashy behavior, which is telling me it's more thread timing related. so now I'm really out of ideas. The DLL still expects the handle to be a pointer type so there's still a mismatch there in the Py file in our add-on though, it's not entirely correct but solving that alone won't fix the weird behavior in headings I'm afraid.
@BTyson yeah it's for sure in that one, just re-installed it though myself to doubly-check. It should say "v0.05" as the version number. If this thing ever makes it to v1.0 stable it'll be a miracle xD
I see it now, cool! This does seem to have somewhat reduced the weird static that can happen, at least with the way I'm running it. If you want to disable phrase prediction I can get you the command to do that too if you want.
When I go into the general dialog box in NVDA and hit enter on Use NVDA during sign-in (requires administrator privileges button, bestSpeak makes a staticky squeak and that dialog crashes and puts me back on the desktop.
Once the NVDA addon is all good to go, all we would need yet is a sapi5 version to use it with balabolka. Balabolka can make sound files of tts engines.
I was an idiot and deleted my post instead of editing it, so apologies if you get this twice. I'm curious how on earth yoiu accomplished this? I see B32_TTS.DLL but have no idea how to begin figuring out what that actually is. Is it some kind of neural voice or did you figure out how to actually emulate the original keynote voice? I thought it was neural until it crashed in a very un-neural-like way.
@jscholes Oh. Well that's pretty cool. I wonder why it took the collective community 30 years to crack this open? I also never even knew the actual name of this until now, so clearly I am living under a rock.
@simon @jscholes so from what I understand: This was found in an old educational product from the mid-90s - the DLL had popped up there thanks to the Internet Archive. Both the 90s software and BeSTSpeech are long dead now, but I find it interesting that the DLL description says: "BeSTspeech TTS for Windows (Designed by Jason Zheng)" and it's copyright 1994. We're using a DLL that's 30 years old, wow, way to celebrate its 30th birthday.
Everybody knows Ubisoft's NFT games and NFT initiative crashed and burned and they had to stop talking about AI, but one other thing I'm noticing is every game with generative AI elements is absolutely slated by consumers - e.g. these are a sample of comments on the Catly trailer from The Game Awards across different channels.
It's exciting to see a new generation just outright reject this stuff.
Want Keynote Gold's Voice inside your NVDA? OK, leaving this here for now. Huge thanks to all of you who jumped in to debug this TTS. Of course, without @rommix0 releasing this find and the demo app, documenting parameters, none of this would have gone anywhere. Oh, and GPT Pro mode still gets credit for teaching me good lessons about threadding exceptions and deadlocks. I tried to add parameters to allow for the reading of full numbers as right now anything above 4-digits gets announced as single numbers. No-go there, n3,0 and n6,0 and n2,0 and n5,0 all just didn't change anything. Maybe someone else can find a fix for that. Most other functions work though. Some people will no doubt ask me to allow this to use another sound output device than default, and that's a big no, it's limited by the DLL itself. So not happening. eurpod.com/BestSpeak.nvda-addo…
Oh my god this is so cool, I'm using this now and I love it! I did notice that it has some delay when reading individual letters, but apart from that its really awesome
so far it's living up to its name with something as this complex for someone who knows Python at such a surface level. I'll also be spending some time at a hot springs for Christmas vacation and rather than having a Be my Eyes volunteer to call all the time to find our way around the various pools, figured we could use the camera with Advanced voice mode. Might as well make full use of it, but I don't think it's something I could pay for every month, only when I'm in higher need for it.
Well, will you please put up some demos. Like show what as do and how you think it works. I know Python at a very surface level myself. I'm just curious more about it.
@alexchapman hmm. Wonder how difficult this would be to solve for, it's interesting that the DLL being loaded in memory and then re-used after switching away to another synthesizer would cause that.
There seems to be an issue where if you, for example, go to run, type "msinfo32," and press enter, when NVDA begins to read that, the speech will start to make weird static for a second, and I'm not sure if that's because of indexing, or because speech is being cut off too quickly. If I cut the speech off quickly in the demo application, this problem doesn't occur. I just wanted to mention this as something to be aware of.
@BTyson was really hoping someone else would do it, and I think it might be that others in the community were making some progress on this, I just opened the pandora's box so to speak for it to have more rapid solutions. But when I want to see something get done I get it done xD
@alexchapman ah darn. I tried to add this function in but it did not solve my problem: ``` def terminate(self): # Called when NVDA switches away from this synth or shuts down self.cancel() # Signal BG thread to end bgQueue.put((None, None, None)) self.bgThread.join()
if self.handle: # Shutup and destroy handle if self.bstShutup: self.bstShutup(self.handle) self.dll.bstDestroy(self.handle) self.handle=None
if self.win: self.win.destroy() self.win=None
# Unload the DLL by removing reference self.dll = None ```
@alexchapman OK! The latest copy up at eurpod.com/BestSpeak.nvda-addo… should gracefully unload the TTS when switching away from it, so hopefully solving a lot of the weird crashes people were seeing at random.
@alexchapman @mcourcel yeah, thankfully I had worked out the default values now with some other folks on Masto, so re-grabbing the file again and updating the add-on should solve any of those quirks :) For sure has that BrailleNote-esk sound!
Okay, this is super cool. Buggy still, but it's trippy to use NVDA and hear KNG on my Windows 11 laptop. It's probably not production-ready just yet, as I'm getting error sounds while typing and it randomly spells all utterences letter by letter. Even so, how incredibly cool!
@vick21 @alexhall KEynote Gold, the repackaged version of Bestspeech that was used in Pulsedata/Humanware products, also offered as a stand-alone/internal synth in both hardware and software form.
How would I add the synthesizer on NVDA? I used to use a Braillenote, I am familiar with Keynote gold. Do you have that for theBraille n -speak synthesizer as well and like the Alex voice or something like that for NVDA
@TwoThousandStu hopefully more experts will jump in :) I think there could be a way to have it recover from that more gracefully rather than taking down the entire main NVDA thread with it, those just are sadly not troubles I'm often facing in the web development world.... xD
@TwoThousandStu I'm even shocked we can run it in modern win32, considering it was built in mid-1995... Shows how at the basics, that compatibility of Windows still shines through. But it also shows just how much things have changed and I feel like in a way you're building fortresses around the unsafe ways and methodologies they used for coding back then to help them adapt to modern times.
@TwoThousandStu yeah, I think what makes this one so frustrating is how the same thing can be read back multiple times, and sometimes it will cause the statick and weird wigging out noises (reminds me of a circuit-bent Furby!) and other times it'll read that same instance of the text just fine. With the way it keeps generating garbage wave data until NVDA fries itself makes me think it's some serious bug in the DLL creating it when it encounters something.
@masonasons @TwoThousandStu headings and links. A good place to sample this and get the bug to show its ugly face is by reading all links on a site like news.google.com - Do you know if perhaps some type of unicode chars get inserted before and after element names by NVDA? That's my best guess for now - it's for sure getting some chunk of text it doesn't like, but then again sometimes it can read it just fine.
@masonasons @TwoThousandStu that's so odd to me, I was able to navigate through the entire freedomscientific.com site's heading structure without it exploding on me, hmm.
@masonasons @TwoThousandStu it could still be thread-management related though, GPT Pro-mode really convinced me of that (+ the suggestions it outlined in eurpod.com/crashes.txt xD) - although the two could certainly be related, in that certain characters cause it to crash or create issues around the bstShutup handling bits.
@masonasons @TwoThousandStu ha, not the energy for hunting down the drivers for them, was going to do that maybe in the next few days and correct any assumptions it's made. Maybe until then someone else will solve where the snag is :)
well, not as much as I would suck just doing it on my own :) while coding all this out the "you've reached the limits of O1" really got to me. And 4O is way, way worse than even O1 is at the coding stuff from my testing, so O1 with Pro mode (and it took 3 minutes and 20 seconds by the way to think that 7-point document out, heh)
@masonasons @TwoThousandStu hmm, I'm not sure though. For me, it's far more productive than learning Python and threadding and all that myself - since I only work on web and docker type stuff for my main job, this is just not at all even in the comfort zone of what I'd be doing, but now thanks to AI getting me started, I still got somewhere. (also it helped me create that Gem simulator web page, but that one I had a lot more hand in and awareness when it was adding some wrong logic.) When I had expressed my frustration with hitting the O1 limits and wanting the pro model nobody talked me out of it, so I got it in haste as well.
@masonasons @TwoThousandStu lol. I also made sure that next month my plan goes back to the plus $20 plan. It's very rare that I would take up a project that blows through all of O1's limits and I don't think that unless they add family sharing to the pro plan the price can be justified. So no, I haven't gone completely off the AI rails, so to speak. :)
@masonasons @TwoThousandStu well not just a keynote driver, also having camera access during my 4-day hot springs vacation, I had other reasons for wanting it :) GPT has helped me learn a lot over time, like when I discovered how mutexes work by working with the C++ interception library - and how to do better thread management. Here, I learned more about BGThread and how that works, about unloading / destroying the DLL properly, and a bunch of other stuff. If I can bite off a smaller project for myself next time then it's quite possible I'll pick up Python not through courses or books but just by making practical examples work with AI.
ha. It thought for another 2 minutes! After I fed it both Espeak and DECTalk's source drivers: "This does not definitively prove that DECTalk must be called only on the main thread or that it can’t be used from a background thread, but the use of a window and messages suggests a design pattern where DECTalk expects a stable, message-driven environment. If calls into DECTalk aren’t synchronized properly or if calls occur on different threads in unexpected ways, crashes can result. This is a heuristic based on experience with similar Windows-based TTS engines and not a guaranteed rule—official documentation or direct testing is the best verification." As for Espeak: "Contrary to what might be inferred from earlier assumptions, NVDA’s eSpeak driver does interact with eSpeak in a threaded environment, but in a way that is officially supported and well-tested. The driver feeds text to eSpeak; eSpeak processes it and calls the callback, and NVDA responds. There’s no manual message pump or requirement that all calls occur on the main thread." Which is I think to say, it reaffirmed itself on point 1, but corrected itself on point 2. Whether with more hallucinations or not? A better Python expert would probs know. xD
I wanted to respond to this original post about the number processing. If you want it to speak numbers as full numbers that are over 4 digits, you have to manually add the commas in the correct places. I think this is why on a BrailleNote, numbers like 1400 would be spoken as "one thousand four hundred," but then numbers like 15400 would be spoken as "fifteen thousand, four hundred."
@J3317 @datajake1999 oh that's odd? Sometimes browsers have changed it to a .zip, but you can always rename the ending extension to .nvda-addon and it should install. (add-ons are just zip packages with a manifest)
If you're able to troubleshoot anything about it, this goes silent if you have multiple profiles, and you switch to a window that uses a tts other than BeSTspeak and then back to one that does.
the other goof I found so far is when adjusting parameters using NVDA synth settings ring, it doesn't speak for every other value adjustment when changing pitch, etc
Wow, this is really neat!!! And I've used ChatGPT to help me with Python stuff. And I might understand less of it than it sounds like you do. I ran the result through a friend and it had only a few errors. Probably right about the speech stuff though, examples help for sure. Thanks for this!
so how does that work? We might be able to grab the ID from the NVDA config (not sure if it expects the string or integer form, either way I guess) and pass it, my guess is you set it before calling TTSWav in the Speak function? Might be doable then to grab the user-set device.
this would be infinitely easier if NVDA 2024.4 returned an integer for the sound device, but still a string until 2025.1 sadly. so then we gotta somehow map convert it, ugh.
just strange. I set it during init of the TTS to 0 or 1, and then it just, falls silent. But then I don't know why, my log showed: "Before TtsWav, GetBstWaveDevID returns: 268538216" which is like tf, unless that's a status code of some sort, it's returning another integer entirely.
sadly, The new logs look similar in terms of the key problem: • The DLL still fails to set the custom device ID (Failed to set wave device ID to 1). • TtsWav returns -5 when trying to speak with that device. This behavior is nearly identical to before—just now we’ve tried different approaches (using unsigned integers, avoiding WAVE_MAPPER calls, using c_void_p for handles), and none have resolved the issue. I'm just sad. frustrated and super sad inside.
Yeah, I switched to using wintypes.UINT instead of c_int for the device ID and made sure the handle is a c_void_p. Unfortunately, even with those changes, bstSetWaveDevID still fails and TtsWav keeps returning -5. Based on your disassembly and testing, it looks like the DLL just won’t successfully open a non-default audio device. But if you ever get this implemented in your dialog reader I could check against that. The code now passes the device ID as an unsigned integer correctly, but the synthesizer still only works when using WAVE_MAPPER (0xFFFFFFFF). It seems the DLL can’t or won’t open other devices, even when given the correct unsigned ID.
aayaya, the bigger problem too is we're doing some magician-converting of the device mappings in NVDA's configurations as strings and trying to map them to correct MME IDs, it's all just added like 3 KB of extra code just to convert the string and try to enumerate them first into a list and then pass the correct one, TBH it's a mess.
have been debugging it for 3 hours now but don't think it'll go anywhere. When 2025.1 comes out and they return integers in the config values for the output device, this could be made a little easier (though I don't know if it will still translate the correct mapping into the MME device ID right), but at least it'll mean not needing to enumerate them based on string name. Right now it's a mess: def enumerateWaveOutDevices(): devices = {} count = waveOutGetNumDevs() for i in range(count): caps = WAVEOUTCAPS() res = waveOutGetDevCapsW(i, byref(caps), ctypes.sizeof(caps)) if res == 0: devName = caps.szPname.strip() devices[devName.lower()] = i return devices and then later: " devices = enumerateWaveOutDevices() log.info("Enumerating devices for exact match:") for dn, did in devices.items(): log.info(f"Device '{dn}' = ID {did}") if val in devices: matched_id = devices [val] log.info(f"Matched exact device '{val}' to ID {matched_id}") return matched_id else: log.info(f"No exact match for '{val}', using WAVE_MAPPER") return WAVE_MAPPER
" So yeah. All because NVDA's config can return a string. The fuck.
nah. MME is getting deprecated in 2025.1 in favor of Wasapi-only.... To them its use is considered legacy unfortunately. And it doesn't help that we cannot pipe the BestSpeech audio directly through Nvwave, because it's another window and handle entirely, it's not as easy as just passing the parameter for the buffers and the dll returns audio to that player object. Gosh my life would have been a dream with Bestspeech if it were that easy.
For anyone who is up for debugging Python code on a Saturday afternoon, please help me get a working copy of BeStSpeak working on NVDA: eurpod.com/BestSpeak.nvda-addo… Notes: - This block the main window thread somehow still, so NVDA thinks it has a dialog open. Consequently you cannot change back to your synth and will need to open NVDA.ini in notepad to change it. - very basic, which means: No rate, pitch, or other synth settings implemented yet.
Interesting. The speech synth window bug specifically affects use of ctrl+nvda+s. But getting to it through nvda settings ends up working, and even allowing you to change synths. However, even when the synth switches back, ctrl+nvda+s is broken for the remainder of the nvda session. Ctrl+insert+s also somewhat works with inconsistent success, however getting to synth settings through nvda settings is consistently successful. Hopefully htis helps to isolate the issue somewhat.
How's this one? dropbox.com/scl/fi/ny40afwm5l2… I just put it in scratchpad alongside the DLL. It has some problems still, for example switching away from it and back sometimes makes it start at a weird samplerate, but it does work.
@TheQuinbox wow, this is fascinating, it's made more with the same threadding model as ESpeak's, admitedly I had a harder time getting how the bgQueue / BgThread class worked, this really helps me. Might as well scrap my version though since this approach is way safer than what I was trying to do.
Thank you to all the ones who submitted a project and wishing you good luck for the next ones. We plan on include more people in the voting process, so stay tuned to hear more about that next year.
Když tedy @archos rozjel českou instanci #PeerTube - vhsky.cz, přemluvil jsem Petra Šimáčka, aby tam dával svá skvělá videa o městské cyklistice. Pomohl jsem mu s přechodem z #YouTube a teď tam najdete vše, co Petr za ten více než rok vytvořil: vhsky.cz/a/mestskacyklistika/
Vím, že se tu o cyklistiku zajímá dost lidí, tak dejte @mestskacyklistika follow, ať Petr vidí, že je tady taky publikum. 😉
Tipy a triky pro městskou cyklistiku. Jsme parta kamarádů, kteří jezdí do práce na kole. Šetří nám to čas, peníze a zlepšuje kondici. Pojďte to zkusit taky a odpoledne už nebudete muset jít do Fitk...
@Milu Já Newpipe nepoužívám, tak asi moc neporádím. Zkusím to odpoledne vyzkoušet. Jinak @mestskacyklistika můžete sledovat i zde na Mastodonu. Federace na vhsky funguje @sesivany
Tipy a triky pro městskou cyklistiku. Jsme parta kamarádů, kteří jezdí do práce na kole. Šetří nám to čas, peníze a zlepšuje kondici. Pojďte to zkusit taky a odpoledne už nebudete muset jít do Fitk...
Asi se to nedá udělat tak, že by se lidi mohli registrovat, aby mohli komentovat, tvořit si playlisty atd, ale prostor pro vkládání videí by měli jen po schválení, že?
@aikencz @Milu @Razemix Jo, lidi, kteří už tu jsou, můžou ke komentování využít svůj existující Fediverse účet, to je super. Jde mi spíš o lidi, kteří se k tomu dostanou odjinud, ale chtěli by zanechat komentář nebo prostě používat jen ten Peertube.
Registrace je potřeba, to je mi jasné. Jde mi o to, že by ty registrace byly otevřené, ale ve výchozím stavu by uživatelé nemohli vkládat žádná videa, jen komentovat, sledovat, dělat si playlisty. A nějakou kapacitu na vkládání videí by dostali až po žádosti a schválení.
@plactagonic Óoooo díky, díky. za nakopnutí. Hned jak mi byl doručen tvůj toot jsem na to přišel. Opět klasická chyba mezi židlí a klávesnicí. Tak teda hlásím, že registrace jsou jíž funkční. 🤦🏽♂️
@Razemix On tam byl nastavený limit registrací na 10 uživatelů a @plactagonic byl přesně 10 uživatel proto už víc nešlo. 🤦🏽♂️🤦🏽♂️🤦🏽♂️ Stane se no 😄 @sesivany @aikencz @Milu
To je ještě ok, já čekal že to bude něco jako vypnuté notifikace nebo nějaké házení do spamu. Já tam stejně začnu něco nahrávat nejdřív za měsíc. @Razemix @sesivany @aikencz @Milu
@Milu @Razemix @schmaker já tedy nevím, ale na mobilu, když rozbalim hamburger, tak mám tlačítko přihlásit a hned pod ním je "žádost o účet". Mám to tak jak ve FF, tak v chomu. 🤷♂️
@Razemix On tam byl nastavený limit registrací na 10 uživatelů a @plactagonic byl přesně 10 uživatel proto už víc nešlo. 🤦🏽♂️🤦🏽♂️🤦🏽♂️ Stane se no 😄 @sesivany @aikencz @Milu
Tuta devs should use AI to help them to fix the UI bugs on Android app. For instance, they should ask on chatgpt: "How to make the date picker in an Android app follow the device mode theme setting?"
Die alte EU-Kommission unter Ursula von der Leyen bekommt wegen DSGVO-Verstößen bei einer Werbekampagne noch einen Verweis des Datenschutzbeauftragten.
Wow! I don't know if folks at the Time realize that they are popularizing such people and add more glamor to their person, no matter their intent. infosec.exchange/@briankrebs/1…
For everyone bent out of shape over Time Magazine naming Trump as the Person of the Year 2024, recall that in 1938 Time featured Hitler as Man of the Year.
I don't claim to have insight on what Time is thinking this time around.
Wie Deutschland bei der Inklusion versagt. andererseits und das ZDF Magazin Royale haben für diese Recherche viele Monate zusammen gearbeitet. andererseits.org/inklusion/
Today is the day I've looked forward too for several weeks, it's the day where I will be joined by many to daybeau the latest album from Christmas Reapers, called "traveling stars." You can be a part of this celibration by tuning to HKC Radio today at 2 PM eastern. hkcradio.com
Musician goes on a link-fest to share his work. That's me, by the way.
Sensitive content
I'm writing a 'Link-fest' post to share with you my various projects in hopes that one of them can lesson your boredom, help with your next presentation or if you like, support what I do. Skip, click or share, it's up to you.
In no particular order then:
1. Free-to-use music for any project containing over a thousand files for hopefully any kind of thing you can think of: onj.me/shorts
2. An even larger collection of longer-form content including the above in it's own folder, plus new covers I've made of my *own* material using #Suno: dropbox.com/scl/fo/jgodxzpzf7s…
3. In September last year I started a podcast with my wife Kirsten called 'StroongeCast' which has it's own weird origins but we're enjoying it very much. AS a blind-sighted couple, we thought we might be a little off the beaten track of most podcasts you listen to, and wanted to share our thoughts with you: 3.1. On youtube we have a dedicated playlist just for the podcast: youtube.com/playlist?list=PLaj… 3.2. The #StroongeCast homepage: onj.me/stroongecast
4. Speaking of youtube, I've been making videos covering various topics since late-2017, which is when I started to take youtube quite seriously. Topics include but are not limited to: 4.1. Main channel link: youtube.com/@TheOnjLouis 4.2. Onj's Tracks. No Talk, All Music: youtube.com/playlist?list=PLaj… 4.3. #InspiredBySound, a series in which I break down some of my tracks, or discuss a sound-library that interests me at the moment: youtube.com/playlist?list=PLaj… 4.4. Onj's Audio and Field Recordings: youtube.com/playlist?list=PLaj… 4.5. A collection of #Suno tracks all based on my own material, not just created with a random prompt. The idea here was to go through my back catalogue to find 30 of some of my best works, and have them recreated/covered by AI. In this way, none of them will hit youtube's content ID and can be used in-the-clear: youtube.com/playlist?list=PLaj…
5. Patreon. I don't mention this in my videos because I leave it at the bottom of video descriptions, always feels a bit dirty to me, so I'll mention it just the once: patreon.com/OnjMusic
6. Youtube membership. Nothing too exciting here and I've tried to keep tears to a small level, but if you rather use this instead of patreon, you can: youtube.com/channel/UCqtMRZeGS…
Thanks so much if you made it this far. I know this has been quite a lengthy read, but in amongst all this, I hope you find something of interest, of benefit or perhaps something you can use in the future, if not right now. #Portfolio
I was just talking to a colleague about the AI bubble. These companies are in so deep they can't tell the truth. They are all lying about the efficacy, costs to consumers and most importantly how & when this tech works or doesn't.
Is there enough money on the line to kill over?
There's likely a trillion bucks of valuations across the industry. Billions in sunk costs, billions in c suite remuneration, billions in VC mgmt costs.
"I found that penguins differed in their reactions to being hoisted between human legs. Some were calm, mildly befuddled at how they got a foot off the ground. Others acted as if they were possessed, squirming and slapping and biting. Penguins are beefy birds, sleek bullets of swimming muscle, torpedoes of power, and they slapped impressively hard."
Live album "XX Years Of Steel" out this Friday!
Order here: https://lnk.to/NoS-XX-YearsOfSteel
Mixed by Christian Ice at Temple of Noise Studios
Video by LyricVids.com
Official HelloWorld Repository: https://github.
Ein juristisches Gutachten klärt, in welchen Fällen digitale Dienstleistungen auch analog angeboten werden müssen. Es kommt zu dem Schluss, dass ein Verbot von Digitalzwang ins Grundgesetz gehört.
@streetcoder "und sich bei einem der großen Gatekeeper für App-Stores registrieren" – genau das! Und noch eins oben drauf: Die beiden sitzen in den USA. Du wirst also quasi gezwungen, Deine PII in ein Land mit einem Datenschutzniveau weit unterhalb der DSGVO zu schicken. Das sollte ein NoGo – und als Zwang selbstverständlich illegal sein.
To everyone finding car safety important, I have some good news and some bad news:
40 out of 45 fatal automated driving crashes reported to the NHTSA and investigated this year were caused by Teslas, but the good news is, this is now being addressed.
The bad news is, the Trump administration will solve this… by removing the [quote:] “excessive reporting” of car accidents.
)
JamminJerry
in reply to Tamas G • • •Tamas G
in reply to JamminJerry • • •JamminJerry
in reply to Tamas G • • •