Search
Items tagged with: AI
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
RNIB Tech Talk: Tech Talk 508: Access AI, WelcoMe In Currys, RNIB Gamer Survey on Apple Podcasts
Show RNIB Tech Talk, Ep Tech Talk 508: Access AI, WelcoMe In Currys, RNIB Gamer Survey - 16 Apr 2024Apple Podcasts
Looking for AI use-cases
We’ve had ChatGPT for 18 months, but what’s it for? What are the use-cases? Why isn’t it useful for everyone, right now? Do Large Language Models become universal tools that can do ‘any’ task, or do we wrap them in single-purpose apps, and build thou…Benedict Evans
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
horse on cohost
fucking, come on man. (edit from the future: a few hours after i posted this, the feature was reportedly turned off - see here [https://cohost.org/nicky/post/5665868-damn-i-should-be-sen] for more info)horse on cohost
"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
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 if you wish investors and marketers would start investing money into HI (Human Intelligence) again.
#HI #HumanIntelligence #AI #ArtificialIntelligence #enshittification
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
GitHub - timpaul/form-extractor-prototype
Contribute to timpaul/form-extractor-prototype development by creating an account on GitHub.GitHub
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
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
Blind internet users struggle with error-prone AI aids
Unreliable software installed to comply with rules to help disabled people navigate online has prompted thousands of lawsuitsFinancial Times
Lots of things happening in the AI/LLM space that could have implications for #accessibility
Ferret-UI from Apple:
https://arxiv.org/abs/2404.05719
ScreenAI from Google
https://research.google/blog/screenai-a-visual-language-model-for-ui-and-visually-situated-language-understanding/
Ferret-UI: Grounded Mobile UI Understanding with Multimodal LLMs
Recent advancements in multimodal large language models (MLLMs) have been noteworthy, yet, these general-domain MLLMs often fall short in their ability to comprehend and interact effectively with user interface (UI) screens.arXiv.org
"Blind internet users struggle with error-prone AI aids", reports an article in the Financial Times:
https://www.ft.com/content/3c877c55-b698-43da-a222-8ae183f53078
#FT has a paywall but three articles can be accessed for free if you register.
TLDR is that AI #accessibility widgets over promise and under deliver. Repeatedly.
#AI #GenerativeAI #a11y
Blind internet users struggle with error-prone AI aids
Unreliable software installed to comply with rules to help disabled people navigate online has prompted thousands of lawsuitsFinancial Times
LOL! So a few blindness higher ups want me to speak at their library events about the use of #AI and how it could enhance the lives of Disabled people, so I go through the song and dance, arranging payment, not at all hiding the fact I am very critical of AI in the disability space because I assume they read my blog by now, but then they come back and say things like, you didn't tell the internet you are actually against AI, to which I send them 5 blog posts where I critically talk about AI and they are always like but we didn't know this!
Pro tip for others. Read my blog, not just social media. That's what my website is for, after all.
Sadly, after they realized I am critical of AI and especially #AIHype they did not get in touch with me to continue the speaking presentation engagement, so this means that no payment changed hands, and I did not get to talk about the more critical aspects that disabled people really should be paying attention to regarding these large language models, such as, the profit driven nature of all of these models. The capital investments that went into these models that are being used in apps like Be my Eyes. The discriminatory side effects of mass deploying these large language models in the workplace, especially in terms of entry-level positions, the privacy and ethical concerns, not to mention, the environmental concerns as well. In short, i’m available for presentations.
New #blog post: MDN’s AI Help and lucid lies.
This article on AI focused on the inherent untrustworthiness of LLMs, and attempts to break down where LLM untrustworthiness comes from. Stay tuned for a follow-up article about AI that focuses on data-scraping and the theory of labor. It’ll examine what makes many forms of generative AI ethically problematic, and the constraints employed by more ethical forms.
Excerpt:
I don’t find the mere existence of LLM dishonesty to be worth blogging about; it’s already well-established. Let’s instead explore one of the inescapable roots of this dishonesty: LLMs exacerbate biases already present in their training data and fail to distinguish between unrelated concepts, creating lucid lies.A lucid lie is a lie that, unlike a hallucination, can be traced directly to content in training data uncritically absorbed by a large language model. MDN’s AI Help is the perfect example.
Originally posted on seirdy.one
: see original. #MDN #AI #LLM #LucidLies
MDN’s AI Help and lucid lies
MDN’s AI Help can’t critically examine training data’s gaps, biases, and unrelated topics. It’s a useful demonstration of LLMs’ uncorrectable lucid lies.Seirdy’s Home
Have you seen @nextcloud AI Assistant 2.0?
You can now chat about your own documents and data, use GPU acceleration, transcribe audio with Whisper, use your own LLMs, and a ton more integrations and new features.
All #open-source 😀
Boosts appreciated 🙏
https://nextcloud.com/blog/nextcloud-releases-assistant-2-0-and-pushes-ai-as-a-service/
#AI #mastodon #Linux #Foss #OSS
Nextcloud releases Assistant 2.0 and pushes AI-as-a-Service - Nextcloud
A major update to the Nextcloud AI Assistant 2.0, plus the news we work with several big hosting providers like IONOS and OVHcloud to bring AI-as-a-Service options to you!Jos Poortvliet (Nextcloud)
https://www.theregister.com/2024/03/28/ai_bots_hallucinate_software_packages/
#security #AI #MachineLearning
AI hallucinates software packages and devs download them – even if potentially poisoned with malware
Simply look out for libraries imagined by ML and make them real, with actual malicious code. No wait, don't do thatThomas Claburn (The Register)
"What made me put my coffee down this morning was 'synthetically generating music', which apparently now falls under 'needs to disclose'.
Them’s fighting words.
Let’s say I fiddle with some knobs on my Eurorack setup (that’s a music synthesizer if you don’t know), and decide to upload it in the background of a video. Am I now making 'altered content?'"
#MusicProduction #MusicTech #AI #Policy #YouTube #Deepfakes
https://vkc.sh/youtubes-bad-ai-disclosure-policy/
YouTube’s bad “AI disclosure” policy – VKC.sh | Veronica Explains
YouTube implemented a new AI disclosure policy. I think it's bad.vkc.sh
#GenAI hype in financial terms! #AI #nvidia #NVDA #generativeAI
https://youtube.com/watch?v=idxW9VGdXoA&si=bYQIKGV8_V3JacDH
The Stock Market is 'Priced to Insanity'.
Is the stock market currently 'priced to insanity'? In this video, we look at what the Dean of Valuation has to say. Aswath Damodaran recently gave an interv...YouTube
Responsible and Ethical AI with Dr. Rumman Chowdhury | Axe-con 2024 Keynote
https://www.youtube.com/watch?v=ccCZ4FxruMU
#ethic #a11y #ai #disability #axecon
Watching this with interest: first #AI (ChatGPT) based screen
reader:https://www.typeahead.ai/
This could change the screen reader paradigm and I'll try the #Windows version when it's available, but I worry that using it for work and remaining #GDPR compliant will not be possible.
This also heralds a much bigger paradigm change - the UI for everyone will change and the GUI as we know it may well disappear.
Typeahead: AI-powered Screen Reader
Typeahead is an AI-powered screen reader for Mac to empower blind and visually impaired users. It uses ChatGPT to navigate through web and desktop apps.www.typeahead.ai
I always enjoy opportunities to learn from Audrey Tang - @au - her recent interview in the podcast #YourUndividedAttention is no exception. In "Future-proofing Democracy In the Age of AI with Audrey Tang" raises important questions about #AI & #Democracy. What do we need to do now to help prepare citizens for #DeepFakes. Important to hear about the role #Taiwan is playing in #ResponsibleAI.
2024 is an election year for almost half of the world's population.
Jakob Nielsen thinks accessibility has failed.
Jakob Nielsen thinks that generative AI will make my experience better.
I think Nielsen needs to think again.
https://tink.uk/nielsen-needs-to-think-again/
#accessibility #a11y #AI
“I’m a Google Cloud software engineer, and I refuse to build technology that powers genocide [and] apartheid”
https://www.dailydot.com/debug/google-protest-israel/
Watch the others in the room. The assholes who jeer and manhandle the protestors. These are your peers in the tech industry. These are the “acceptable folks” in tech. Fuck these people.
Look at their faces. They know exactly what they’re complict in. Look at their smug smiles. Dickheads each one.
#tech #google #ai #ProjectNimbus #israel #genocide #palestine
Google Protest At Tech Conference Over Israel Contracts
Google engineers protested Google at an Israel tech conference in New York before being thrown out by security.Marlon Ettinger (The Daily Dot)
https://t.ly/fO2WU
#PerceivePossibilities #Accessibility #AI #ArtificialIntelligence #LetsEnvision #AssistiveTechnology
Envision App 3.3 Launches 'Ask Envision' for Scanned Text & A Richer 'Describe Scene'
Every update is a step towards greater accessibility and empowerment.t.ly