As #Israel uses #US-made #AI models in #war, concerns arise about tech’s role in who lives & who dies
"an #AP #investigation revealed new details of how #AIsystems select targets & ways they can go wrong, incl'g faulty data or #flawed #algorithms.. based on internal docs, data & exclusive interviews w curr & fmr #Israeli #officials & co employees.. is the 1st confirmation tt #commercialAI models r directly used in #warfare.. enabling tis type of unethical & #unlawful war”
https://apnews.com/article/israel-palestinians-ai-technology-737bc17af7b03e98c29cec4e15d0f108
#AISystems
0 posts0 participants0 posts today
"Even the most impressive #AIchatbots require thousands of human work hours to behave in a way their creators want them to, and even then they do it unreliably. The work can be brutal and upsetting, as we will hear this week when the ACM Conference on #Fairness, #Accountability, and #Transparency (#FAccT) gets underway. It’s a conference that brings together research on things I like to write about, such as how to make #AISystems more #accountable and #ethical." https://www.technologyreview.com/2023/06/13/1074560/we-are-all-ais-free-data-workers/
MIT Technology ReviewWe are all AI’s free data workers
Replied in thread
(4/n)
...in many #EU languages here: (3).
If you live in the #EU, you are lucky that there seem to be quite a number of #MEP's that have enough know-how to ascertain risks that the #European #Commission had originally missed in the proposition:
"MEPs substantially amended the list to include bans on intrusive and discriminatory uses of #AISystems such as:
“Real-time” remote #biometric identification systems in publicly accessible spaces;...
(3)
https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52021PC0206
eur-lex.europa.euEUR-Lex - 52021PC0206 - EN - EUR-Lex
Is anybody else getting tired of the idea that AI is somehow "alive" or will be soon? Let's have a reality check.
Rabbit Trails and Parting VeilsAI: Automated Insanity!
Replied in thread
@ShadSterling @simon @annaleen @BBCWorld
(6/n)
"...was because of the "*unexpected* acceleration" in #AISystems that "we *need* to take a step back".
Holy...the #LLM's have already reached a level where they grow *inexplicably* faster!
If you apply #NiklasLuhmann, another #godfather, but of the #SociologicalSystemsTheory, to #GenerativeArtificialInteligence, you could construct the following hypothesis:
I guess we can all easily agree that #AGI's are #complex systems.
If you...
As part of pre-release safety testing for its new #GPT4 #AI model, launched Tuesday, #OpenAI allowed an AI testing group to assess the potential #risks of the model's #emergent #capabilities—including #PowerSeeking #behavior #SelfReplication, and #SelfImprovement. While the testing group found that GPT-4 was "ineffective at the autonomous replication task," the nature of the experiments raises eye-opening questions about the safety of #future #AISystems.https://arstechnica.com/information-technology/2023/03/openai-checked-to-see-whether-gpt-4-could-take-over-the-world/ #AI #Risk #safety
Ars TechnicaOpenAI checked to see whether GPT-4 could take over the world"ARC's evaluation has much lower probability of leading to an AI takeover than the deployment itself."
The whole conversation about #AGI tickles me.
What we have in #AI, #ML is a gamut of computational, intelligent tools and some empirical evidence about what their strengths and weaknesses are.
To solve ANY problem, we need insightful #AI #Systems #AISystems research - how do we put the tools together into something that works for a specific problem, measured through careful #experiments.
But, alas, this isn't as exciting as #AGI - which remains undefined and triggers ideological sparring.