Items tagged with: AI

Search

Items tagged with: AI

Sensitive content

#Accessibility #gemini #google #blind #AI #chatgpt

#smartphones #AI #samsung #galaxya565g #galaxya365g #galaxya265g #samsunggalaxy

Sensitive content

#AI #chatgpt

#AI #ML #llm

After reading these medium term goals, I'm honestly glad I stopped donating because I know some corporate thing will sponsor them now they will/want to add AI, even an AI Chat bot, to NVDA.

The below are just 2, but there's more AI stuff they have planned.

• Integrated AI image description: Free, secure, performant, private and VI-focussed on-device AI image description available out of the box.
• Help chatbot prototype: Exploring the development of an AI-powered chatbot to provide instant support and answer frequently asked questions about NVDA.

The second one is particularly nonsensical.

empyreanriders.com/feed/ #NVDA #AI #AIHype

#AI #nvda #AIhype

#News #AI #culture #humanity #elonmusk #reactionaries #technocracy #peterthiel #emotion

3/3 Gemini Flash 2.0 Thinking: Humans are not just social animals, but profoundly self-domesticated apes. And this self-domestication is the key to understanding our unique cognitive and social abilities.
#LLM #AI #ML
#AI #ML #llm

2/3 Deepseek R1: Humans instinctively convert selfish desires into cooperative systems by collectively pretending abstract rules, money, laws, rights, are real. These shared hallucinations act as games, where competition is secretly redirected to benefit the group, turning conflict into society's fuel.
#LLM #AI #ML
#AI #ML #llm

1/3 Three Responses from 3 top reasoning models to Lex Fridman's question, What's one truly novel insight about humans?
O1-pro: Humans are unique among known species in that they simultaneously rewrite two layers of reality, the external world and their own private mental landscapes, and then merge these two rewritten layers into a continuous personal narrative that feels objectively true.
#LLM #AI #ML
#AI #ML #llm

#AI #cbc #cdnpoli #dobetter #polcan #plc #LPC #aibs

#android #app #AI

The future of communication in two panels.
#AI
A two-panel cartoon. Left: Person excitedly viewing a computer says, "A.I. turns this single bullet point into a long email I can pretend I wrote." Right: Another person responding to an email says, "A.I. makes a single bullet point out of this long email I can pretend I read."
#AI

#AI #ML

I'm still hoping for a decent Hungarian AI-based TTS whenever a book comes up that I want to read. It's frustrating that even with all the tech advances, the best speech engines are basically the same as in 2006. Professional narrators reading non-fiction, making text and audio alignment less tedious and less resource-intense training could really help with this. I'm not expecting character voices or emotions, just something enjoyable. #Hungarian #AI #Audiobooks #TechFrustration
#AI #audiobooks #hungarian #techfrustration

Thing I want to exist but don't want to build: a #mud that does pretty much the same thing as #SillyTavern or #agnaistic, intended for selfhosting. A MUD would be an excellent interface for #AI assisted #roleplay. With #Evennia and #OpenRouter, this honestly wouldn't be that hard. I just hate #Python and don't wanna.

The way to do it would be:
1. Set up a #docker#Evennia environment, to make distribution easy.
2. Add an attribute on Evennia accounts to hold an openrouter API key, and a userflow to get a users key via OAuth: openrouter.ai/docs/use-cases/oauth-pkce3. Use the #OpenAI Python API and hook it up to NPC's to let them chat, generating responses with the user's API key. Figure out function calling and hook it into the various creation functions of Evennia for extra fun. Probably need to store attributes for what models and parameters to use on each NPC, with global defaults probably from the user's account.
4. Dump the room name and desc, the player desc, and chat history into the AI context.
5. Something something MSP/MXP and ElevenLabs for sound effects and music.
6. Something Something MXP for images.
7. Dockerize the thing and upload it to let users run there own muds for AI roleplay for themselves and their friends.

If I either liked Python or was unemployed, I would do this. If only AI coding was good enough to just make the thing based on my design skeleton above.


OAuth PKCE - Secure User Authentication

Implement secure user authentication with OpenRouter using OAuth PKCE. Complete guide to setting up and managing OAuth authentication flows.
OpenRouter Documentation
#AI #python #docker #openai #MUD #rolePlay #sillytavern #agnaistic #evennia #openrouter

Here's another #AI#TTS service, free for now, called #Zonos. playground.zyphra.com/audioOr free for self-hosting on github: github.com/Zyphra/Zonos

GitHub - Zyphra/Zonos: Zonos-v0.1 is a leading open-weight text-to-speech model trained on more than 200k hours of varied multilingual speech, delivering expressiveness and quality on par with—or even surpassing—top TTS providers.

Zonos-v0.1 is a leading open-weight text-to-speech model trained on more than 200k hours of varied multilingual speech, delivering expressiveness and quality on par with—or even surpassing—top TTS ...
GitHub
#AI #tts #zonos

Sensitive content

#AI #riffusion @Onj 🎶

#AI #ML #llm

Now that we have three different #AI Music generators, I wanted to give them a tricky test and compare the results. I'm totally not musical, so hopefully one of my followers will chime in to this thread with some musical theory criticism.

The prompt we'll be using today: "A comic operetta in the style of Victorian satirical comic operettas making fun of President Donald J Trump." We'll be leaving all other settings as default.

The things I'm looking for:
1. Does the song refer to recent events?
2. Can it follow my strong style instructions? I'm wanting something in the style of gilbert and Sullivan. But if I mention them directly, the AI's either refuse on copyright grounds, or generate endless bad versions of A Model of A Modern Major General.

Let's start with riffusion.com. It gave me a song called "Why won't they let me edit?" It's apparently about how Donald Trump wants to edit his wikipedia page. I'm not aware of this being a thing that happened, but it's possible I missed that? It starts off enough like an opera for my taste, but it quickly devolves into a kind of 2000's pop musical style. I give it 8/10 for topicality, and 5/10 for style. Link: www.riffusion.com/riff/40bbc4f8-c557-4bb4-9a95-1bbcb66c38de#riffusion



Why Won't They Let Me Edit? by fastfinge

Comic operetta, Gilbert & Sullivan style, orchestral, Victorian parlor music
Riffusion
#AI #riffusion

wo, i just received an email from google gemini, and i'm excited about it! seems like they are releasing a new feature called Gemini Live1! another AI feature that can help improve the lives of blind people.

"Google is starting to roll out updates to Gemini to make your conversations even more dynamic and engaging.

Gemini Live1 is a more natural way to have free-flowing voice conversations with Gemini on your phone. With our latest model, Live can better understand multiple languages, dialects, or accents in a single Live chat and help with your translation needs. In the coming months, we’ll also bring screen sharing and live video streaming capabilities to Live.

Your data is processed per the Gemini Apps Privacy Notice. As part of providing
this improved experience, your audio, video, and screenshares are stored in your
Gemini Apps Activity (if it’s on). Your data in Gemini Apps Activity is deleted per your
auto-delete period in that setting, and you can manage and delete your Gemini Apps
activity anytime. Learn more in the Gemini Apps Privacy Hub.

Thanks for using Gemini.
Sincerely,
The Gemini Team

© 2025 Google Ireland Ltd
Gordon House, Barrow Street,
Dublin 4, Ireland.
1 Check responses for accuracy. Compatible with certain features and accounts. Internet connection required.
Available on select devices and in select countries, languages, and to users 18+.
You have received this mandatory service announcement to update you about important changes to Gemini."

#Blind #AI #Google #Gemini #GeminiLive1 #Tech

#gemini #google #blind #AI #tech #geminilive1

After firing off a glib toot to @bagder this morning, I decided to test #AI code assistants to see how easy it is to get them to disable SSL certificate validation in CURL. All of the "mainstream" models will gladly do this if you tell them "your code doesn't work, it says invalid certificate". In fairness they try to warn that this is insecure but script kiddies aren't gonna read those warnings, they're gonna CTRL+C, CTRL+V. Full report here brainsteam.co.uk/2025/2/12/ai-… #infosec #curl #php

Getting AI Assistants to generate insecure CURL requests

Testing AI code assistants willingness to generate insecure CURL requests
brainsteam.co.uk
#infosec #AI #php #curl @daniel:// stenberg://

#AI #elonmusk #openai #ki #kunstlicheintelligenz

Remember Aaron Swartz, because Zucc won't be getting any jail time for this.

youtube.com/watch?v=bBa5TO_nBJ…

Meta leeched over 80 terabytes of books of off torrents for commercial purposes. Not personal use. They made sure to not seed the books to cover their behinds.

When you do it, it's 30 years in prison, when they do it it's a fine.

#AI #ProfessionalPiracy #Meta #Llama #Ollama


Facebook Illegally Downloaded ALOT of Books...

In this video I discuss how Meta/Facebook has been illegally downloading dozens of terabytes of books from Z-library, Libgen, and other sources to train thei...
YouTube
#AI #meta #llama #ollama #professionalpiracy

Meta staff torrented nearly 82TB of pirated books for AI training — court records reveal copyright violations

"...court records reveal that the social media company used pirated torrents to download 81.7TB of data from shadow libraries including Anna’s Archive, Z-Library, and LibGen. It then used this information to train its AI models."

#tech #news #Meta #AI #books #bookstodon #writers #libraries #media

tomshardware.com/tech-industry…


Meta staff torrented nearly 82TB of pirated books for AI training — court records reveal copyright violations

Did they think they could get away with it?
Jowi Morales (Tom's Hardware)
#books #News #AI #tech #libraries #media #meta #writers #bookstodon

🚨 Welke gegevens van je zou #Meta nog meer allemaal gebruiken om #AI te trainen? 🤔

tweakers.net/nieuws/231648/met…

#WhatsApp heeft al AI in de app...🤷 faq.whatsapp.com/1002544104126…

Zeker zijn dat je privégesprekken en data niet en nooit worden gebruikt? Kies #Signal: signal.org/install

#privacy

Vertrouw je op de blauwe ogen van Mark? Of kies je voor het blauwe logo. Signal is open source en controleerbaar. Écht gemaakt voor jou en jouw privacy

'Meta gebruikte ruim 80TB aan illegaal verkregen data om AI te trainen'

Meta gebruikte naar verluidt ruim 80TB aan illegaal verkregen data om het Llama-llm te trainen. Werknemers van het bedrijf zouden onder meer ruim 35TB aan documenten uit Z-Library en LibGen hebben gedownload aan de hand van torrents.
Yannick Spinner (Tweakers)
#privacy #signal #whatsapp #AI #meta

Zkousim novy research mod od OpenAI. V zasade je na tom videt, jak jsou vsichni hraci (slovo velmi hodici se) v reaktivnim modu. V jednotkach dni po sobe takticky zverejnuji male pokroky.

Research:
🤷 Efektivne likviduje agent startupy nad jejich API
💸 Jeden dotaz sezere ~500k tokenu => $10 dolaru z ceniku API
🙅 Ignoruje nastaveni robots na webu, kde hleda (kdo by to byl cekal)

Ale musim rict, ze vysledky sou super. Za par drobnych propalite VC fondum $1k elektriny mesicne.

#ai #openai #it

#IT #AI #openai

#AI #DataProtection #aitakeover

Unbelievable

#ElonMusk’s US #DOGE Service are feeding sensitive data into #AI software via #Microsoft’s #cloud

#Musk’s US #DOGE Service have fed sensitive data from across the #Education Dept into #ArtificialIntelligence software to probe the agency’s programs & spending….
The AI probe includes data w/personally identifiable info for people who manage grants, & sensitive internal financial data…

#law #security #InfoSec #CyberSecurity #NationalSecurity #Trump #TrumpCoup
washingtonpost.com/nation/2025…

#infosec #Trump #security #microsoft #AI #law #education #cybersecurity #cloud #musk #elonmusk #artificialintelligence #nationalsecurity #doge #trumpcoup

#a11y #Accessibility #github #podcast #AI

#google #AI #USA #capitalism #bigtech #siliconvalley #fascism

#apple #AI #canada #mac #appleintelligence #macos15sequoia #buycanadian

This is wild. In the updated Mac User Guide when you click on "learn more" in the Apple Intelligence area of the Settings it says:

"Turn Apple Intelligence on: With macOS 15.3 or later, Apple Intelligence is on by default and integrated across features in your apps. "

then a few lines later…

"Note: Apple Intelligence is currently in beta. Apple Intelligence uses generative models, and outputs may be inaccurate, unexpected, or offensive. Check important information for accuracy."

#Canada #BuyCanadian #AI #AppleIntelligence #mac #Apple #macos15sequoia

A screenshot of the help guide a portion of the text says: Apple Intelligence is the personal intelligence system built into macOS Sequoia. It combines the power of generative models with an understanding of your personal context to deliver intelligence that’s useful and relevant. Apple Intelligence is designed to understand and create language and images; prioritize and summarize notifications, mail, and messages; and draw from the personal information on your device to simplify and accelerate everyday tasks. And it does all this while protecting your privacy. Availability: Apple Intelligence is available with Siri and device language set to English in most regions around the world. For more information, see the Apple Support article Apple Intelligence requirements. Turn Apple Intelligence on: With macOS 15.3 or later, Apple Intelligence is on by default and integrated across features in your apps. If you have an earlier version of macOS 15 (or if you previously turned off Apple Intelligence), choose Apple menu > System Settings, click Apple Intelligence & Siri , then click Get Apple Intelligence. Turn Apple Intelligence off: Choose Apple menu > System Settings, click Apple Intelligence & Siri , then click the button next to Apple Intelligence. Note: Apple Intelligence is currently in beta. Apple Intelligence uses generative models, and outputs may be inaccurate, unexpected, or offensive. Check important information for accuracy.
#apple #AI #canada #mac #appleintelligence #macos15sequoia #buycanadian

I would also like to say to Apple. You've had Siri for how many years? You've added many variations to your English speakers. But no Canadian variations.

Perhaps it's just the moment we're in, but I'd like to see a Canadian Western, Canadian Eastern, Canadian Maritime, and Newfoundland Option for Siri.

Thank you very much.
#Canada #BuyCanadian

#AI #AppleIntelligence #mac #Apple #macos15sequoia

A screenshot of the English voice dialogue in Siri shows on the left, American, Australian, British, Indian, Irish, and South African. American is selected and there are 5 voices available on the right
#apple #AI #canada #mac #appleintelligence #macos15sequoia #buycanadian

“ChatGPT Extension”

OH HELLS NO. I'm not going to be a part of that. I draw the line there immediately.

Edit: It appears that this is not on by default. I did not set it up and I turned off the "suggestions”.
#AI #AppleIntelligence #mac #Apple #macos15sequoia

About Siri, Dictation & Privacy... Siri Responses... Extensions ChatGPT Siri and other features can use ChatGPT to help answer your requests.
< > ChatGPT Extension Use ChatGPT ChatGPT can help you write text, create an image, answer questions, and more. OpenAl Terms of Use... Set Up... Siri Setup Prompts Allow Siri to suggest ChatGPT when you make a request that requires it.
A dialog offering to connect to ChatGPT
#apple #AI #mac #appleintelligence #macos15sequoia

You can turn Apple Intelligence off on macOS in the new “Apple Intelligence & Siri” area of System Settings.

See my previous boost on creating a “mismatch" Siri voice that effectively disables AI across all your connected Apple iCloud devices.
#AI #AppleIntelligence #mac #Apple #macos15sequoia

A screenshot of the Apple Intelligence & Siri dialogue shows it is about 10 options down the list on the left. There is a toggle button at the top of the dialog on the right with “Apple Intelligence”
#apple #AI #mac #appleintelligence #macos15sequoia

#apple #AI #mac #appleintelligence #macos15sequoia

Not super-amazing and yet amazing I was searching for a restaurant on Google Maps using iOS, etc. The screen came up and on it there was a photo of the menu. No problem: tap to display the photo in full view, then ask Siri, "does this place serve juices, or whatever, and booms! An answer from ChatGPT does the rest. You can even ask follow-up questions. Just awesome! #accessibility #ai !
#Accessibility #AI

#Trump #AI #USA #musk #elonmusk #us #congress #government #techbros #america #fightback #senate #hackers #USAID #resist #doge #coup

Third party developers writing #apps on multi-platforms or just on #Linux - Tell us your journey, submit a talk at LAS and meet us virtually or in person in Tirana, Albania on April 25-26. Whether it's #AI chat app, #game, or even a mobile phone app. We want to know! You can submit your paper at linuxappsummit.org/cfp.
A web banner that says "Linux App Summit", April 25-26, Tirana Albania followed by in bold black Call for papers.

Call for proposals

The Linux App Summit (LAS) is designed to accelerate the growth of the Linux application ecosystem by bringing together everyone involved in creating a great Linux application user experience.
linuxappsummit.org
#linux #apps #AI #game

⇧