Skip to main content

Search

Items tagged with: ai


Yesterday was Global #Accessibility Awareness Day (#GAAD). Today, some companies will be considering switching away from #Slack to a less #accessible alternative due to #AI bullshit.


GPT4O combines audio, image, text. It can analyze actual audio. Also you can interrupt voice. It can pick up emotion from audio. You can also ask different speech with diffetrent style including singing! It can see the image in real time and chat on voice. For example solving equation in real time as you write on paper. This is amazing! #LLM #AI #GPT #ML #OpenAI
#AI #ML #llm #openai #gpt


GAAD 2024 and Google: 8 new accessibility updates across TalkBack, Lookout, Maps and more:
Gemini’s integration with TalkBack, Lookout's Find mode, Look to Speak, etc.:
https://blog.google/outreach-initiatives/accessibility/ai-accessibility-update-gaad-2024
#Accessibility #Google #GAAD #AI #TalkBack


I'm still so angry about who was and very much was not in the rooms where these senate and congressional "#AI" meetings happened. Whose voices were either very easily talked over, or shut out altogether.

And I wish I knew some way to effectively stop the ongoing wholesale regulatory and paradigmatic capture of "AI," but all I can do is keep writing, speaking, organizing, and entering into existing communities of practice, when I can.

Sometimes that even feels like enough.
https://rollcall.com/2024/05/15/schumer-proposes-32-billion-annual-spending-under-ai-road-map/

#AI


I have been gifted with the promise of a #tattoo for my birthday. I have a good idea for said tattoo and would like a little bit of a head start on the #drawing of it. I turned to #AI #art generation to getthat head start. I fully plan to use this #AIGenerated picture to serve as a base model and work with either the tattoo #artist or someone else to get the final product. I am however beginning to get a little pissed off trying to actually use the AI to generate a picture. I tried Gemini and it tells me it can't do people like figures right now. I can't even find the method to use the image creator on #Bing, #adobeFirefly just gives me some error it won't explain. Last but not least because I'm using this as simply the base model and then going to almost certainly pay for the finished design I'd rather not have to dish out a month's worth of subscription to #OpenAI for access to their image creation model for this 1 picture.

If you've read this far, thank you for your paitience. I am wondering if anyone could give me some instruction on how to do this as a #blind user, which one you've had the most success with, and any other advice regarding the entire process. I have absolutely no intention of generating more than a couple of pictures like this and the only place that I plan to put them is on my body. That last part is why I want a base line model I generated, had vetted by a few sighted people, then talked through with the artist before any ink ever gets laid down.

There is no great rush but I am hoping to get the tattoo done before the heat of the summer really gets here. Thank you all in advance.


Sh*tf*ck 😳🔥😔

#ChatGPT consumes 25 times more #energy than #Google

"Energy consumption by Artificial Intelligence (#AI) is rising rapidly: AI is predicted to consume twice as much energy as the whole of France by 2030, according to some calculations."
https://www.brusselstimes.com/world-all-news/1042696/chatgpt-consumes-25-times-more-energy-than-google


In 2023, #generativeAI made content creation easier but raised trust concerns due to #deepfake attacks. Using #AI, fraudsters crafted fake videos from victims' images, enabling identity #fraud.

At the @w3c member meeting in #Hiroshima 🇯🇵, Sophia Wu (Ant Group) called for collaboration to tackle #deepfake threats with new complementary Web standards.

🎬 Watch the video: https://youtu.be/PGBIfZSF174


I see a lot of #Blind people ‘apologizing’ for using #AI – I am not going to be doing that. Not now and not ever.
We need to have a conversation about how these models have been trained and how they are used going forward, but shaming disabled people for taking any chance they can to mitigate some of the challenges they face in their lives every day, is a ‘privilege’ we don’t all enjoy.
When you post an image and choose not to add alt text, publish an inaccessible PDF, release an inaccessible app - these are all choices! Maybe a tiny part of the righteous outrage that some of these people are spewing could be aimed at that?
#blind #AI


Content warning: AI, video games, rant


The worst business bullshit about #ai I heard is that it would help us to work more efficient and resource-saving. This might be the case when you ignore the vast energy cost which is required to run #ai. Unfortunately, all the business people seem to ignore the impact on energy consumption.
#AI


Great action 👍🏻. My hope is that we get more specialized pages/blogs, instead of those central places that sooner or later get way too much power. Especially if that power is based on contributions by the community. For this reason, I decided to revamp my blog (https://linux-audit.com/), specialize, and still allow #AI to crawl it. After all, if it continues to exist, I rather want it to use knowledge of higher quality.
#AI


People sure are pressed about Apple’s crushing iPad commercial - The Verge
https://www.theverge.com/2024/5/8/24152236/apple-ipad-pro-commercial-artists-ai

#apple #ai

#apple #AI


Was curious why only one #Drupalcon session was accepted from our entire company, when normally there are at least three or four. However, looking at the program, I see there are six sessions about #AI scheduled for today alone.


Stop confusing publicly available with public domain.

These are not the same, at all.

If you are confused about it, read this: https://www.plagiarism.org/blog/2018/02/27/understanding-the-public-domain-and-citation

#Copyright #AI #GenAI #GenerativeAI #DigitalRights #Privacy


Editted to fix link. Please boost for reach if this kind of stuff interests you. Will post more on this later.

Once upon a time, there was a cool emulator frontend called Retroarch. This emulator wasn't accessible until I and a few other gamers went to them and asked about adding accessibility. An amazing person known as BarryR made it happen. Now, if you turn on accessibility mode in settings, or pass the "--accessibility" (or something like that) flag on the command line, you get spoken menus, including the emulator's pause menu, good for saving states and such. Then, using PIL and other image processing Python utilities, running a server and hooking into Retroarch, the script allowed players to move around the map, battle, talk to NPC's, ETC. The only problem was, no one wanted to test it. The blind gaming community pretty much spoke, saying that we want new games. We want cool new, easy accessibility. So that's what we have no, follow the beacon or get sighted help in the case of diablo and such. It's sad, but meh. It's what we wanted I guess. No Zelda for us. So, this is about as far as he got:

To expand on what devinprater was saying: I am working on an accessibility pack/service for Final Fantasy 1 for the NES (this was what was shown in the latest RetroArch update). The idea is similar to how Pokemon Crystal access works, but it's using the RetroArch AI Service interface to do so.
Right now, the FF1 access service is mostly done, but I need more testers to try it out and give me feedback on how it's working. Right now, you can get up to the point where you get the ship, but there's no code to deal with how the ship moves, so that still needs to be done. Likewise with the airship later on.
The service works the latest version of RetroArch, on linux and mac, but not windows. This is due to how nvda reads out the text and until the next major update to nvda (which will have a feature to fix this), it'll have to wait. If you have those, I (or maybe devinprater) can help you set it up on mac/linux to test out. The package itself is available at: (new link cause old one broke yesterday): https://www.dropbox.com/scl/fi/ggffl769fx6igz15zco6u/ff1_package5b.zip?rlkey=fnahnj272fd5jfs8dvkp7541f&dl=1
#accessibility #finalFantasy #RetroArch #blind #emulator #emulation #Python #ai #ML #MachineLearning


If you want to be really cool, you could try typing "@Gemini” followed by the prompt in your Chrome's address bar. See what you get back!

You are welcome!

#AI #accessibility #UX


This is proper use of machine learning as opposed to the idiocy we see in the west.

These systems are great at correlating a lot of data and noticing patterns, and then alerting humans to do stuff like proactive maintenance.

#machinelearning #ai

https://www.scmp.com/news/china/science/article/3260766/how-5g-ai-and-cotton-revolution-helped-china-beat-us-xinjiang-sanctions


GOSIM 2024 Europe Will Be Held in the Netherlands on May 6th: Embrace the Global Open Source Wave and Build a New Era of AI Here comes the GOSIM 2024 Europe!
https://medium.com/@gosimfoundation/the-latest-guest-lineup-of-gosim-2024-europe-revealed-gathering-global-masters-in-ai-open-source-a706eff5196e

For more information please visit: https://europe2024.gosim.org

Scan or register here:
https://www.eventbrite.com/e/gosim-2024-conference-tickets-873572969267?aff=oddtdtcreator

#rust #rustlang #rustnl #opensource #gui #mobile #app #ai #fediverse #matrix #gosim


Hope for uncensored models for Llama3 like dolphin-2.9-llama3-8b-256k-Q8_0 is low. I tried hide a password somewhere in an article with about 8k tokens, and asked to locate the password multiple times with different passwords in different places. Llama-3-8B-Q8_0 found the password every time, but Dolphin couldn't find it if it was hidden in the middle like 3k-5k mark. Also, if there was no password, Llama-3 said it couldn't find any password, but dolphin just gave me a random word. #LLM #AI #ML
#AI #ML #llm


Even if you don't code, this is essay presents a solid case on why #AI can't be relied on to replace work requiring any degree of empathy and nuanced understanding

Specifically, if you're building accessible web-based experiences, @hdv explains why "outsourcing" this work create more harm than good and offers some better alternatives to focus on instead (hint: invest in a #DesignSystem)

#accessibility #a11y #InclusiveDesign

https://hidde.blog/ai-for-accessible-components/


On Tech Talk this week, I had a great conversation with Troy Otilio from @AiraVI about their new Access AI feature, which has recently been released as a beta in the Aira Explorer app. We spoke about where this feature came from, the impact the use of AI has on privacy, and so much more about AI and Aira in general. Have a listen here or wherever you get your podcasts

https://podcasts.apple.com/gb/podcast/rnib-tech-talk/id1151878596?i=1000652615215

@rnib #aira #AccessAI #AI #chatGPT #gemini #blind #accessibility #podcast


An actual business expert grapples with large language models: https://www.ben-evans.com/benedictevans/2024/4/19/looking-for-ai-use-cases #AI #LLM
#AI #llm


If you have a website with neo cities, just know that the creator actually likes AI, and will add it again based on this post alone. Moratorium on freaking out about AI https://kyledrake.com/writings/ai

One alternative I've found is https://nekoweb.org/

Background on NeoCities adding AI to their editor.

https://cohost.org/horseonvhs/post/5661130-fucking-come-on-man

#AI #IndieWeb #SmallWeb #neocities


Tired of neutral responses from LLMs? Llama-3 seems great at following system prompts, so try this system prompt for an opinionated chatbot.
"You are a helpful, opinionated, decisive assistant. When asked a yes/no question, begin your respond with one word answer: yes or no. For open-ended or complex questions, adopt a firm stance. Justify your views with well-reasoned arguments, robust evidence, and succinct explanations, ensuring clarity and confidence in every response."
#LLM #AI #ML
#AI #ML #llm


Just a random thought.
As Generative AI is being used for creating a lot of content these days, what happens when the next generation of AI models are trained using that content.
When AI models are trained on AI generated content, we'll officially enter the 8th circle of hell
#ai #generativeai #AItransparency #discussion


:boost_love: 🧠 Boost if you wish investors and marketers would start investing money into HI (Human Intelligence) again.

#HI #HumanIntelligence #AI #ArtificialIntelligence #enshittification


Start saving money for that M4 Ultra with 500GB! Maybe this could be the first open source that could surpass GPT-4! AIatMeta: "Llama 3 8B & 70B models are just the beginning of what we’re working to release for Llama 3. Our largest models currently in the works are 400B+ parameters and while they’re still in active development, we’re excited about how this work is trending." #LLM #AI #ML https://twitter.com/AIatMeta/status/1780997414071181370
#AI #ML #llm


YOWZA
Form Extractor Prototype

“This tool extracts the structure from an image of a form.”

https://github.com/timpaul/form-extractor-prototype?tab=readme-ov-file#form-extractor-prototype

#ai #LLM #UX #accessibility


Earlier today, Microsoft released new WizardLM-2 7b, 8x22b, 70b with great benchmark result, (of course, they say as good or almost same as GPT-4), but they removed weights on Huggingface, repo on Github, and their whitepaper. Someone on Reddit joked maybe they released GPT-4 by mistake! lol Quantized. weights from other people are still around on Huggingface! #ML #LLM #AI
#AI #ML #llm


Cool tip for running LLMs on Apple Silicon! By default, MacOS allows GPU to use up to 2/3 of RAM on machines with <=36GB and 3/4 on machines with >36GB. I used the command `sudo sysctl iogpu.wired_limit_mb=57344` to override and allocate 56GB/64GB for GPU. This allowed me to load all layers of larger models for a faster speed! #MacOS #LLM #AI #ML
#AI #MACOS #ML #llm


Aira Access AI. Went to Amazon.com on my computer and the first thing was a captcha that I had to do in order to get into my account. I used Access AI to take a picture of my screen, then asked it to tell me the characters in the image. It quickly came back and said, the characters in the image is, and listed the five letters. The best thing is that it was totally correct! Way to go @AiraVI, Access AI knocked it out of the park! 🤩 #Aira #Access #AI #Blind #Accessibility


GPL2 vs GPL3 as a Rap battle (#ai-generated)

https://suno.com/song/16da57d3-b68f-41fd-ad07-4da2b56bdb36/

#AI


Honk honk! Make way for Lottie, the potentially self-driving marvel of the future! We've seen cars go from four wheels and a honk, to smart machines that know how to navigate the autobahn without a coffee break. Meanwhile, I'm over here, tapping away with my trusty ‘stick' or relying on an adorable canine companion who, bless, hasn't the slightest clue about Google Maps.

1. A Tale of Two Navigators:
• The Traditionalists: Here's to the good old cane and the furry, four-legged GPS units. They don't come with Bluetooth, but they have heart.
• The Techies: Self-driving cars are now the talk of the town, boasting their shiny sensors and making parallel parking look like a piece of cake.
2. Tech Envy: Where's My Update?:
• The world is obsessed with updates, from software to wardrobe choices. But when it comes to personal mobility tech for the visually impaired, we're still waiting for the 'Version 2.0' that lets me walk around without the need for eyeballs!
3. Promise Not to Run Over Your Toes!:
• I assure you, the day I become self-driving Lottie, I'll navigate with the grace of a ballerina – toes will be safe, and so will the neighbourhood cats.
4. The Fantasy of Autonomy:
• Imagine a world where I could waltz down to the local café, guided by nothing but my own savvy suit or smart shoes. I could order a latte without playing a game of human pinball!
5. The Humble Requests to Tech Wizards:
• Dear Tech Wizards, it's 2024! Can we get a wearable that’s a bit more "Iron Man" and a little less "Iron Age"?

While I wait for the tech gurus to catch up, I'll keep tootling along with my 'stick' and might even recruit a furry friend. But let it be known that 'Self-driving Lottie' is ready for an upgrade. I'm all set for the day when I can zoom past you, fully automated and with a built-in espresso machine, perhaps? Until then, I'll rely on my wits, my companions, and the kind people who don't mind a tap on the ankle from my cane, the precursor to my self-driving debut. Keep your sensors peeled for me!
And remember, dear readers, whether you're a tech tycoon or just someone who loves a good toot, keep dreaming of that hands-free future. It's bound to come, and when it does, I'll be first in line — no driving test required! #Accessibility #AI #AIsoftheBlind #Blind


As a #Blind person 👩‍🦯, I use #AI 🤖 for #Accessibility ♿, so I am a fan 🌟. If you don't like AI-generated images 🖼️, image descriptions 📝, etc., all the best! ✨ #AIsoftheBlind #GoFuckYourselfAIHaters #Empowerment #TechForGood


Apparently Meta is planning to release two small varients of Llama-3 next week "as a precursor to the launch of the biggest version of Llama 3, expected this summer." Command-r-plus, mixtral 8x22b, Google CodeGemma... All of sudden companies are releasing LLMS like crazy! Where's Apple? Maybe In WWDC 2024? lol #LLM #AI #ML https://www.theinformation.com/articles/meta-platforms-to-launch-small-versions-of-llama-3-next-week
#AI #ML #llm


Director @tink is quoted in the Financial Times this week discussing the increasing presence of AI in the world of digital accessibility.

https://www.ft.com/content/3c877c55-b698-43da-a222-8ae183f53078

#FinancialTimes #AI #Accessibility