#trustworthy ai

[ltsglobaldigitalservices]

Understanding Trustworthy AI: Essential Components and Practices

Trustworthy AI pertains to the creation and utilization of AI systems that function dependably, ethically, and with transparency. It must adhere to principles concerning fairness, accountability, integrity, privacy, and safety within AI applications and decision-making protocols. These fundamental principles serve as the foundation for ensuring responsible AI usage and reducing potential harm. A recent notable instance involves the proliferation of Generative AI for content creation, which has underscored challenges in managing deepfake content and addressing bias within AI-generated outputs.

For detailed information, visit us to see the full article here!

[aitalksblog]

Building Safe AI: Anthropic's Quest to Unlock the Secrets of LLMs

(Images made by author with Microsoft Copilot) Large language models (LLMs) like Claude Sonnet are powerful tools, but their inner workings remain shrouded in mystery. This lack of transparency makes it difficult to trust their outputs and ensure their safety. In this blog post, we’ll explore how researchers at Anthropic have made a significant contribution to AI transparency by peering inside…

[clarkjason]

The image classifer is in agreement about this treachery. This is a coffeepot.

As part of my research informatics work, I've been prototyping how we can use #machinelearning to create metadata or connect it to scholarship and research processes. This image classifier uses #transferlearning based on the ImageNet database of 14,197,122 images (http://image-net.org) and samples a few @msulibrary digital collection images. Use the link below and refresh to cycle through the predictions.

https://www.jasonclark.info/files/image-classifier/

Opening René Magritte jokes aside... This prototyping has been instructive to understand the limitations of #machinelearning image classification. More importantly, it gives me faith that this can be improved with some additional models. The #TensorFlow library I'm using here allows me to register the confidence it has in the predictions. Moreover, I can train and serve an additional model to help it. There is a path to improvement.

Too often, I've seen these limitations put forward as reasons to not pursue or worry about how #machinelearning will work in cultural heritage settings. A refrain of "Oh, that dumb machine can't figure out what this is!" But, the scale and opportunity to refine our metadata and classification routines within these tools is too hard for me to ignore. The scale gives me pause. Because if we get classification and description wrong, we not only impact discovery, we might also label someone's humanity incorrectly. See image below where Native American context is not present in the prediction.

These #machinelearning classification applications need to be grounded in requirements of "do no harm" & ethics. Sara Mannheimer has been leading some of our local and national work around these ethical questions and I'm looking forward to seeing where we go from here. And if you are interested, code for the #transferlearning image classifier is available here: https://github.com/jasonclark/image-classifier.

[razzek]

A quip about AI I was thinking about yesterday:

AI ingests human input (words, art, etc) much the same way a jet engine ingests flocks of birds, breaking everything down into a slurry of it’s former parts. And much like an engine ingesting a bird, everything falls out of the sky as a result.

[alatariel-galadriel]

professor requires use of generative AI to come up with research questions, 10 dead 15 injured

[remix-of-your-guts]

It wasn't a random urinal, duchamp made it, something ai can never do

arguably the most important fact about Fountain was that duchamp very much did not make it. nearly every art historian agrees it was a random urinal from a plumbing store and the entire point of the piece was that it was a mundane object that he did not make, elevated to art by the act of simply calling it art. it was about showing that art is not just about literal creation but more importantly about the creation of ideas. is performance art not art? is photography not art? there absolutely exist forms of art where the artist did not craft anything with their hands. i'd argue the fundamental thread behind all art is that it is about conveying ideas, whether that idea is aesthetic, practical, philosophical, absurdist, or anything else in nature. and that was kind of the entire point of the dada movement, to challenge existing notions of what "art" is.

i don't necessarily agree with the dadaists on everything- i have great respect for art as a concept and a cultural keystone, but that is also the reason why i have to acknowledge that anything put forth by humans for no other purpose than to exist and to express a feeling, a message, an idea, is absolutely art and is the thing (in my opinion) that sets us about from animals. so yeah, sometimes that includes ai generated images lmao (even if i have IMMENSE issues with the datamining and anti-worker practices that have quickly taken over nearly all of current ai).

[queenbread]

I'm not completely sure how I got to watching pirated spongebob

Is this even SCP anymore

[barbex]

I saw a tiktok today that was obviously entirely made by AI:

The grandma, the voice. It's nearly perfect, there's just a bit of flickering on the hat/shawl, and her movements are a bit too smooth. The voice is also really good, very smooth, missing just a bit of emotional emphasis.

Maybe I'm just too old but I find this horrifying. The nice grandma makes this feel so friendly and personal, grandma would never lie, would she? But none of it is real. She's just made to appear trustworthy by an algorithm.

[paterday]

Posting this over here cuz I wanna show friends my wips but I don’t wanna be annoying about it

[ai-chatter]

Let's kick off with this one because a. it talks about an educational role (I'm not convinced, though I could see the same argument used for search engines so maybe) and b. because it hits the main discussion point in a lot of current articles - that current chatbots are very good at sounding authoritative but don't have any safeguard for not giving false info. More on that soon..

[aitalksblog]

Ethical AI: The Foundation for a Brighter Digital Future

(Images made by author with Leonardo Ai) The increasing use of AI systems in everyday life raises a host of ethical questions and challenges, such as bias, privacy, and transparency. This blog post examines the concept of ethical AI, synthesizing crucial insights from recent research and guidelines. By embracing an ethical framework that takes these guidelines into account, AI stakeholders can…

[gothasmo]

If Ais potentially finds MC at the beginning and brings them to Kuras, I wonder if Kuras directs patients that are beyond his abilities to the Seaspring as a last resort as their final choice...

[aceredshirt13]

well this is the first and strangest spam comment I’ve ever gotten on a fic.

extremely bold of this bot to assume an AI could write my stupid-ass Ritchieverse oneshot in inaccurate screenplay format about Holmes not wanting to take a bath

anyway screen your comments folks

[grisklein]

NOT THE AI THERAPIST DIAGNOSISING ME WITH OCD

[angelofsmalldeath-codeine]

[rockpapercynic]

The internet is flooded with AI-generated photos and they're getting harder to spot.

Most of the time AI is used for click-farming, but lately images have been used in fake news stories and product scams.

Most important: THINK CRITICALLY. AI will eventually get too good to make obvious mistakes. Being media literate means checking not just if an image is real, but if the source is trustworthy.

If you're not sure, don't share! You might be spreading misinformation.