‘#TED is ground zero of the #AI gold rush. But there was also cheerleading and l’ve been overwhelmed by huge love and support from others who see exactly what is happening. It’s the weirdest time to be here.
And it was the weirdest energy from an audience of any talk I’ve ever given. But then, it was intended to make them uncomfortable.
Politics is technology now.
#SiliconValley is desperate to deny that, but it can’t and nor can we.’
Carole #Cadwalladr
The European Union has launched its AI Continent Action Plan, aiming to build AI gigafactories and bolster computing infrastructure
#AI #EU #Europe #EuropeanUnion #Gigafactory #AIFactory #Datacenters #AIInfrastructure #AIRace #InvestAI #EuroHPC #AIAct
A new report reveals much of the internal turmoil behind Apple Intelligence's revamped version of Siri. Read more at @macrumors. #AI #Apple #AppleIntelligence #Siri #Tech #Technology https://flip.it/zA1bpz
Samantha Enslen hires writers, editors, and designers, so I am especially interested in what she has to say about AI and how it's changing work.
Check out the whole conversation:
Watch: https://youtu.be/Hs7kV26AQNo
Read: https://grammar-girl.simplecast.com/episodes/ai-future/transcript
If a company can skew its scam "AI" to give answers it wants but not answers it doesn't then AI is about as reliable a source of information as its' owner is.
Would you trust Elon Musk, Sam Altman or Mark Zuckerberg to tell you the truth or give a balanced account of a subject?
"AI" is the invention of lying machines.
The last few days I am heavily thinking about Memetics which is an extremely cool idea.
And now LLM Agents become really widespread with new standardization tools and will find their way in _a lot_ of tech we will soon perceive as totally normal.
So I think there will appear memetic viruses (kinda like promt injections) that will infect the current context “of” a model and replicate themselves via websites, data stores, social media posts and general output. There will be semi model agnostic memetic viruses.
It is just a matter of time until they appear in a widespread fashion.
This will kickstart a small industry that will collect vectors of these memetic viruses, kinda like anti virus software made their money in the past.
All this will not work to 100% and there will be eventually an memetic virus that will bypass all these measures. It is more likely it appears randomly then being engineered.
This will lead to widespread awareness of memetic viruses like general propaganda and memetic warfare what will lead to the public demanding to criminalize social media platforms to help to distribute memetic viruses.
This therefore will lead to a full circle back to a non globalized world of values, ideas and concepts, like it was 100 years ago.
Memetic viruses like “it’s ok to be white” or “the world is flat!” that spread over continents and the world, will be contained into their local subcultures or zones.
Political extremism will die down until only highly localized, small groups of individuals will be affected for strong non common perspectives.
The global culture will not have a chance to develop and the world will, yet again, develop strong local cultural values and rituals.
Ethnologists will have a bright future.
Europe unveils plan to become ‘AI continent’ with simpler rules, more infrastructure
https://zurl.co/7u779
#ai #airegulation #eu
I’m fed up of hearing ‘if an #AI doesn’t produce a good response it’s because the prompt was no good’. If the system is incapable of producing the correct output in response to a simple and obvious prompt, it isn’t fit for purpose. Plus, since you already have to know what the correct answer is in order to judge whether the generated output is any good, there’s little point in using #generativeAI in the first place.
In an early March announcement that was underreported at the time (but just became a lot more relevant in early April) the Trump regime revealed its intention to collect and presumably monitor social media handles of folks applying to legally work and reside in the United States:
https://theintercept.com/2025/03/23/trump-immigrants-social-media-citizenship-green-card/
Trump Wants Immigrants on U.S. Soil to Hand Over Social Media Accounts to Apply for Citizenship
"Collecting social media information, according to the USCIS proposal first posted March 5, is necessary “for the enhanced identity verification, vetting and national security screening.”
The proposal specifically cites Trump’s January 20 executive order, which advocates have warned goes well beyond the Muslim travel ban from Trump’s first term, which targeted people living abroad.
The new executive order stated that “the United States must ensure that admitted aliens and aliens otherwise already present in the United States do not bear hostile attitudes toward its citizens, culture, government, institutions, or founding principles, and do not advocate for, aid, or support designated foreign terrorists and other threats to our national security.”
USCIS said the social media handles it collects would be used to determine if people applying for a variety of immigration statuses pose a “security or public-safety threat.”"
Of course, observers immediately noted the connection between this proposed policy and Trump's ideological policing, kidnaping, and attempted deportation of foreign students who participated in anti-Genocide protests and/or oppose the Trump regime:
"In light of Columbia University protester Mahmoud Khalil’s ongoing detention, one official from a Muslim civil rights group said the new policy poses special danger for critics of Israel and the Trump administration.
“This policy would disparately impact Muslim and Arab applicants seeking U.S. citizenship that have voiced support for Palestinian human rights,” said Robert McCaw, director of government affairs at the Council on American-Islamic Relations. “Collecting the social media identifiers of any potential green card applicants or citizens is the means to silencing their lawful speech.”
Furthermore, as a representative of the EFF presciently observed, there are few if any limits on the scope of what the government can do with this data if they get it, and modern AI technology allows for the regime to ideologically police and target a terrifying number of people who oppose our fascist overlords. It's also, strictly speaking, a violation of Constitutional rights:
"The policy proposal does not sketch out limits on how USCIS can use its newly acquired data, according to Saira Hussain, a senior staff attorney at the Electronic Frontier Foundation.
Hussain said she was particularly concerned that the government might use artificial intelligence or other automated tools to punish speech it dislikes, pointing to a news report that the State Department is using AI to revoke the visas of people who allegedly express “pro-Hamas” sentiments.
Hussain said she feared a chilling effect, where people applying for a change in status refrain from speaking about potentially controversial issues.
“Anybody who is within the bounds of the United States has First Amendment rights,” she said. “The Constitution applies whether you are somebody who is a citizen or somebody who is a green card holder who is here in the United States. I think that this administration is trying to chip away at that notion, but that is very much what First Amendment jurisprudence has been under the courts.”
Finally, it should be pointed out that the regime's proposal gives no indication whatsoever of when the government would stop recording and monitoring the social media activities of the people it's targeting:
"CAIR’s McCaw said he worried that the policy could be used to continue tracking people’s activity on social media even after they become naturalized citizens.
“There’s no clear sign on when this intrusion into our electronics and communications will end,” he said."
Scientists created Shark AI, a free curriculum that lets students build computer vision models to classify fossil shark teeth, helping teachers integrate AI into science lessons in a fun way.
https://theconversation.com/shark-ai-uses-fossil-shark-teeth-to-get-middle-school-kids-interested-in-paleontology-and-computer-vision-251125
#education #artificialintelligence #AI
#sharks #paleontology #AI #teaching #DigitalLiteracy
"We are researchers in AI literacy and STEM education who helped create a series of lessons that use fossil shark teeth to demonstrate the power and pitfalls of AI. The curriculum guides middle school students and teachers through building and evaluating computer vision models that can reliably classify fossil shark teeth."
Large Language Models are Unreliable for Cyber Threat Intelligence.
This isn’t actually surprising, we can even generalise it: Large Language Models are unreliable. It’s a property, and that’s totally fine, you just need to be aware of it.
Just added: Cats Taking Naps sticker set!
High-quality, original illustrations, perfect for decorating planners, scrapbooks, and gifts.
Order now: https://curiousgoblin.etsy.com/listing/1897733903
#CatsOfMastodon #Cat #Cats #Mastocats #catstodon #fedicats #caturday
#CatStickers #StickerAddict #Stationery #Sticker #Stickers #StickerArt #StickerCommunity
#Art #Mastoart #BuyIntoArt #AIArt #AI #Illustration #Etsy #EtsyShop #Scrapbooking #Scrapbook #Mushroom #Fungi
#AI
It is inevitable that accelerationists will push for acceleration and that the acceleration will generate extremely abrupt deceleration
Do you use AI?
█▓▒░ 
Mia’s Simulacrum 
![Au bureau, un manager l’air satisfait balance dans un sourire : « Lefebvre, votre poste a été remplacé par une IA ! » Lefebvre s’offusque : « Impossible ! » Puis ils se tournent tous deux vers le poste mal éclairé de Lefebvre, où un robot déprime devant un écran d’ordinateur en lâchant un « Je veux crever ! » mécanique. Lefebvre ne peut cacher sa stupéfaction : « C’est effrayant comme ça progresse vite. » [fin]](https://media.oisaur.com/media_attachments/files/114/312/776/294/799/744/original/82adb5333fff4c27.png "Au bureau, un manager l’air satisfait balance dans un sourire : « Lefebvre, votre poste a été remplacé par une IA ! » Lefebvre s’offusque : « Impossible ! » Puis ils se tournent tous deux vers le poste mal éclairé de Lefebvre, où un robot déprime devant un écran d’ordinateur en lâchant un « Je veux crever ! » mécanique. Lefebvre ne peut cacher sa stupéfaction : « C’est effrayant comme ça progresse vite. » [fin]")