@Whats_AI @the2ndfloorguy Happy to help. These systems and those behind them need to think ethically from the get-go.
English
AIAAIC
1K posts
AIAAIC
@AiControversy
Collecting, dissecting and surfacing #AI, #algorithmic and #automation incidents and controversies. Making the case for #transparency #openness #accountability.
Cambridge, UK; Global Katılım Aralık 2020
576 Takip Edilen564 Takipçiler
@AiControversy @the2ndfloorguy Good to see these being tracked systematically. The pace of new AI incidents is a bit alarming TBH.
English
Thank you for sharing this @the2ndfloorguy.
We've summarised this episode so that people can better understand the ethical issues and harms raised by the use and misuse of AI. It's a great example.
aiaaic.org/aiaaic-reposit…
Pankaj@the2ndfloorguy
my cook was stealing fruits from my fridge so i deployed my ai roommate in the kitchen and it caught her red handed 👩🍳 it monitors the kitchen when she's cooking and pings me the moment she takes anything. sends me weekly report. i also trained it to check if she washes her hands before cooking and cleans the slabs once done. I caught her twice this week. just fired her next i'll add gas detection and idle time tracking. a lot more to build. tbh it's not a great version, it's early and rough but yet but it's already catching things i never would have noticed. fn, using claude haiku 4.5 as the vision model. moving it to a local model on my raspi soon so it runs fully offline. btw i run face detection locally and blur it before hitting llm
English
Tracking "AI" harms, resource.
#AIAAIC incident report on #CrisisTextLine fm Jan 2022.
The AIAAIC site & undertaking are impressive. "AI, Algorithmic and Automation Incidents and Controversies" @AiControversy
aiaaic.org/aiaaic-reposit…
English
I think Sora 2 is over. After a few great days, the fun has died down, the best posters have left, and I can't convince any of my friends to sign up
English
@adamdangelo Its not just inaccuracies - it's fun and games and deliberate misuse that is driving a great deal of sloppiness. And its just so easy to do. For which, congratulations!
English
I think "slop" is a great term in the current moment, but I also expect it will become dated within a few years as AI improves and its output generally gets to fewer inaccuracies than the human-created material it is compared against.
English
Here's a breakdown of the risks & harms, and transparency and accountability limitations of OpenAI's Sora models.
aiaaic.org/aiaaic-reposit…
#sora #copyright #disinformation #impersonation #IP #bias #privacy #AIethics
English
Seeing @sama turned into a deepfake supermarket thief is just the start of the genAI video slippery slope - and we're not just talking his extremtly laissiez-faire approach to copyright.
#sora #copyright #disinformation #impersonation #IP #bias #privacy #AIethics
English
We’re working towards making Lovable the most secure place to build software.
These are some of the security improvements we’ve recently shipped:
- Detecting RLS incorrect usage
- Deep code security review
- Warnings if you put an API key in the chat
[More details in thread]
English
I was replaced by AI at my job and have been officially unemployed since last week. Are we still claiming that AI is incredible and changing humanity for the better? It’s soulless. It’s slaughtering humanity.
English
Much trumpeted @DefraGovUK/@NaturalEngland
UK AI-generated peatland map is found to be highly inaccurate, with potentially profound implications for policymaking, environmental degradation
aiaaic.org/aiaaic-reposit……
#AI #AIethics #environment #UK
English
@jason_kint Guillaume Lample left Facebook to found Macron's beloved Mistral AI, and you can bet they have been up to the exact same trickery - aka theft
English
It's time to call it. AI is built on a house of cards of intellectual property violations starting with Facebook which is starting to look a lot like a crime scene as held back discovery documents begin to be compelled and unsealed in court. /1
English
We are collecting examples of actual and potential harms caused directly or indirectly by #DeepSeek
Feel free to share investigations, assessments, audits, legal dockets, media coverage, etc here:
aiaaic.org/aiaaic-reposit…
English
Here's a dispassionate look at the benefits, risks and harms of Midjourney's AI image generator
#midjourney @midjourney #AI #AIethics
English
Incident AIAAIC1830: Top AI firms, incl @AnthropicAI @Apple @business @Meta appropriate 100,000+ scripts to improve AI system dialogue outputs
aiaaic.org/aiaaic-reposit…
h/t @_alexreisner
#AI #AIethics #copyright
English
AIAAIC retweetledi
AI has become a black box in more ways than one. In an excerpt from "Taming Silicon Valley," @GaryMarcus recommends citizens demand transparency from #AI companies around several key factors: bit.ly/3XU0gFa via @bigthink
English