#ai update

[aiupdates]

🌐 Welcome to the forefront of artificial intelligence innovation AdeptAI

Progressive ML research and product lab, committed to shaping the future of general intelligence.

[8pxl]

PSA: Tumblr/Wordpress is preparing to start selling our user data to Midjourney and OpenAI.

you have to MANUALLY opt out of it as well.

to opt out on desktop, click your blog ➡️ blog settings ➡️ scroll til you see visibility options and it’ll be the last option to toggle.

to opt out on mobile, click your blog ➡️ scroll then click visibility ➡️ toggle opt out option.

if you’ve already opted out of showing up in google searches, it’s preselected for you. if you don’t have the option available, update your app or close your browser/refresh a few times. important to note you also have to opt out for each blog you own separately, so if you’d like to prevent AI scraping your blog i’d really recommend taking the time to opt out. (source)

[talktikitaca]

6월 5일 AI Gemini 제미니 업데이트 예정

다음 Gemini 업데이트는 이번주 수요일 2024.06.05 예정되어 있는 것으로 보입니다.

[saragrosie]

Sketching while streaming s5...

Jonathan Sims I will learn to draw you (this is my doing. I could draw him however I want and I choose to stick with an image of him in my brain that is difficult for me to draw. Masochism.)

Not s5 Mahtins below I enjoyed drawing cuz hes neat:

(Edit: I yassified Martin in the do not separate cuz I wanted his hair fluffier)

[dutechsolution]

Enhancing Gmail Security: Google's Latest AI Update Unveiled Google has unveiled a groundbreaking update to Gmail's spam detection, reinforcing email security. Dive into the future of safer online communication as Google employs cutting-edge technology to combat spam effectively. Explore the article for an in-depth look at this game-changing update, ensuring your Gmail experience is more secure than ever. Stay informed, stay secure. Learn more: https://dutechsolution.com/google-revealed-new-ai-update-to-make-gmail-secure-spam-have-a-look/ 

[sayruq]

A little-discussed detail in the Lavender AI article is that Israel is killing people based on being in the same Whatsapp group [1] as a suspected militant [2]. Where are they getting this data? Is WhatsApp sharing it? Lavender is Israel's system of "pre-crime" [3] - they use AI to guess who to kill in Gaza, and then bomb them when they're at home, along with their entire family. (Obscenely, they call this program "Where's Daddy"). One input to the AI is whether you're in a WhatsApp group with a suspected member of Hamas. There's a lot wrong with this - I'm in plenty of WhatsApp groups with strangers, neighbours, and in the carnage in Gaza you bet people are making groups to connect. But the part I want to focus on is whether they get this information from Meta. Meta has been promoting WhatsApp as a "private" social network, including "end-to-end" encryption of messages. Providing this data as input for Lavender undermines their claim that WhatsApp is a private messaging app. It is beyond obscene and makes Meta complicit in Israel's killings of "pre-crime" targets and their families, in violation of International Humanitarian Law and Meta's publicly stated commitment to human rights. No social network should be providing this sort of information about its users to countries engaging in "pre-crime".

[nobodysuspectsthebutterfly]

FYI artists and writers: some info regarding tumblr's new "third-party sharing" (aka selling your content to OpenAI and Midjourney)

You may have already seen the post by @staff regarding third-party sharing and how to opt out. You may have also already seen various news articles discussing the matter.

But here's a little further clarity re some questions I had, and you may too. Caveat: Not all of this is on official tumblr pages, so it's possible things may change.

(1) "I heard they already have access to my data and it doesn't really matter if I opt out"

From the 404 article:

A new FAQ section we reviewed is titled “What happens when you opt out?” states “If you opt out from the start, we will block crawlers from accessing your content by adding your site on a disallowed list. If you change your mind later, we also plan to update any partners about people who newly opt-out and ask that their content be removed from past sources and future training.”

So please, go click that opt-out button.

(2) Some future user: "I've been away from tumblr for months, and I just heard about all this. I didn't opt out before, so does it make a difference anymore?"

Another internal document shows that, on February 23, an employee asked in a staff-only thread, “Do we have assurances that if a user opts out of their data being shared with third parties that our existing data partners will be notified of such a change and remove their data?” Andrew Spittle, Automattic’s head of AI replied: “We will notify existing partners on a regular basis about anyone who's opted out since the last time we provided a list. I want this to be an ongoing process where we regularly advocate for past content to be excluded based on current preferences. We will ask that content be deleted and removed from any future training runs. I believe partners will honor this based on our conversations with them to this point. I don't think they gain much overall by retaining it.”

It should make a difference! Go click that button.

(3) "I opted out, but my art posts have been reblogged by so many people, and I don't know if they all opted out. What does that mean for my stuff?"

This answer is actually on the support page for the toggle:

This option will prevent your blog's content, even when reblogged, from being shared with our licensed network of content and research partners, including those that train AI models.

And some further clarification by the COO and a product manager:

zingring: A couple people from work have reached out to let me know that yes, it applies to reblogs of "don't scrape" content. If you opt out, your content is opted out, even in reblog form. cyle: yep, for reblogs, we're taking it so far as "if anybody in the reblog trail has opted out, all of the content in that reblog will be opted out", when a reblog could be scraped/shared.

So not only your reblogged posts, but anyone who contributed in a reblog (such as posts where someone has been inspired to draw fanart of the OP) will presumably be protected by your opt-out. (A good reason to opt out even if you yourself are not a creator.)

Furthermore, if you the OP were offline and didn't know about the opt-out, if someone contributed to a reblog and they are opted out, then your original work is also protected. (Which makes it very tempting to contribute "scrapeable content" now whenever I reblog from an abandoned/disused blog...)

(4) "What about deleted blogs? They can't opt out!"

I was told by someone (not official) that he read "deleted blogs are all opted-out by default". However, he didn't recall the source, and I can't find it, so I can't guarantee that info. If I get more details - like if/when tumblr puts up that FAQ as reported in the 404 article - I will add it here as soon as I can.

Edit, tumblr has updated their help page for the option to opt-out of third-party sharing! It now states:

The content which will not be shared with our licensed network of content and research partners, including those that train AI models, includes: • Posts and reblogs of posts from blogs who have enabled the "Prevent third-party sharing" option. • Posts and reblogs of posts from deleted blogs. • Posts and reblogs of posts from password-protected blogs. • Posts and reblogs of posts from explicit blogs. • Posts and reblogs of posts from suspended/deactivated blogs. • Private posts. • Drafts. • Messages. • Asks and submissions which have not been publicly posted. • Post+ subscriber-only posts. • Explicit posts.

So no need to worry about your old deleted blogs that still have reblogs floating around. *\o/*

But for your existing blogs, please use the opt out option. And a reminder of how to opt out, under the cut:

The opt-out toggle is in Blog Settings, and please note you need to do it for each one of your blogs / sideblogs.

On dashboard, the toggle is at https://www.tumblr.com/settings/blog/blogname [replace "blogname" as applicable] down by Visibility:

For mobile, you need the most recent update of the app. (Android version 33.4.1.100, iOs version 33.4.) Then go to your blog tab (the little person icon), and then the gear icon for Settings, then click Visibility.

Again, if you have a sideblog, go back to the blog tab, switch to it, and go to settings again. Repeat as necessary.

If you do not have access to the newest version of the app for whatever reason, you can also log into tumblr in your mobile browser. Same URL as per desktop above, same location.

Note you do not need to change settings in both desktop and the app, just one is fine.

I hope this helps!

[four-eyed-floozy]

Some shitty transparents of the anniversary gaku and gumi. Yeah, it's got shitty upscaling, but I really wanted to see their outfits, and I didn't want to wait for the full pngs so I got lazy

[timetravellinghipster]

9 Month AI Update - Twerking is banned but boy can he salsa

A star is born

Yas queen

Oh she thiccccc

[aiupdates]

With Agoge AI, users can expect a tailored approach to their communication needs. From perfecting negotiation techniques to enhancing presentation skills, this AI tool provides training to excel in business communication. 🗣️👥

[the-lambda-archives-ai]

[1/?]

Next > here

Prev > here

[blueberryfruitbat]

AI and Tumblr

In the event of any kind of AI Art API is to be trained on this platform I do not give permission for my art to be sampled or used to train AI without financial compensation, artists hold the copyright to our art and any intentional mimicry or sampling of my art by an artificial intelligence is infringing on an artist's rights to their creations.

We as artists will not take this grim proposal between Tumblr and Midjourney laying down, this was the last safe heaven for a lot of us and you are now holding knives to us entirely to solve your financial woes instead of offering meaningful solutions. You are going to bleed your community of all trust and all revenue you still eek out of us. Your ways have changed from one of a community built around passion for fandom and the creative mind to one of greed and exclusion. You have been showing true colors and they are ugly, they are sickening, and they will kill your platform. If this deal goes through, may your platform die slow and painfully, may you watch your most loyal of whales walk away and leave you drying out in the hot pavement that is bankruptcy.

[bogkeep]

there's something so deeply dystopian to me how tech companies don't understand that a forced convenience is not a convenience at all. i'm sure autocorrect is helpful for many, but a function that forcibly changes my actual written words and punctuation is taking away my language. photo filters can be nice but i need to choose using them myself or else i have lost the ability to take the picture i want. i don't want a machine to draw or write for me. taking away the option for me to do things manually feels like violence!!!! all this talk of endless opportunity, why are you RESTRICTING me

[mostlysignssomeportents]

Google is (still) losing the spam wars to zombie news-brands

I'm touring my new, nationally bestselling novel The Bezzle! Catch me TONIGHT (May 3) in CALGARY, then TOMORROW (May 4) in VANCOUVER, then onto Tartu, Estonia, and beyond!

Even Google admits – grudgingly – that it is losing the spam wars. The explosive proliferation of botshit has supercharged the sleazy "search engine optimization" business, such that results to common queries are 50% Google ads to spam sites, and 50% links to spam sites that tricked Google into a high rank (without paying for an ad):

https://developers.google.com/search/blog/2024/03/core-update-spam-policies#site-reputation

It's nice that Google has finally stopped gaslighting the rest of us with claims that its search was still the same bedrock utility that so many of us relied upon as a key piece of internet infrastructure. This not only feels wildly wrong, it is empirically, provably false:

https://downloads.webis.de/publications/papers/bevendorff_2024a.pdf

Not only that, but we know why Google search sucks. Memos released as part of the DOJ's antitrust case against Google reveal that the company deliberately chose to worsen search quality to increase the number of queries you'd have to make (and the number of ads you'd have to see) to find a decent result:

https://pluralistic.net/2024/04/24/naming-names/#prabhakar-raghavan

Google's antitrust case turns on the idea that the company bought its way to dominance, spending the some of the billions it extracted from advertisers and publishers to buy the default position on every platform, so that no one ever tried another search engine, which meant that no one would invest in another search engine, either.

Google's tacit defense is that its monopoly billions only incidentally fund these kind of anticompetitive deals. Mostly, Google says, it uses its billions to build the greatest search engine, ad platform, mobile OS, etc that the public could dream of. Only a company as big as Google (says Google) can afford to fund the R&D and security to keep its platform useful for the rest of us.

That's the "monopolistic bargain" – let the monopolist become a dictator, and they will be a benevolent dictator. Shriven of "wasteful competition," the monopolist can split their profits with the public by funding public goods and the public interest.

Google has clearly reneged on that bargain. A company experiencing the dramatic security failures and declining quality should be pouring everything it has to righting the ship. Instead, Google repeatedly blew tens of billions of dollars on stock buybacks while doing mass layoffs:

https://pluralistic.net/2024/02/21/im-feeling-unlucky/#not-up-to-the-task

Those layoffs have now reached the company's "core" teams, even as its core services continue to decay:

https://qz.com/google-is-laying-off-hundreds-as-it-moves-core-jobs-abr-1851449528

(Google's antitrust trial was shrouded in secrecy, thanks to the judge's deference to the company's insistence on confidentiality. The case is moving along though, and warrants your continued attention:)

https://www.thebignewsletter.com/p/the-2-trillion-secret-trial-against

Google wormed its way into so many corners of our lives that its enshittification keeps erupting in odd places, like ordering takeout food:

https://pluralistic.net/2023/02/24/passive-income/#swiss-cheese-security

Back in February, Housefresh – a rigorous review site for home air purifiers – published a viral, damning account of how Google had allowed itself to be overrun by spammers who purport to provide reviews of air purifiers, but who do little to no testing and often employ AI chatbots to write automated garbage:

https://housefresh.com/david-vs-digital-goliaths/

In the months since, Housefresh's Gisele Navarro has continued to fight for the survival of her high-quality air purifier review site, and has received many tips from insiders at the spam-farms and Google, all of which she recounts in a followup essay:

https://housefresh.com/how-google-decimated-housefresh/

One of the worst offenders in spam wars is Dotdash Meredith, a content-farm that "publishes" multiple websites that recycle parts of each others' content in order to climb to the top search slots for lucrative product review spots, which can be monetized via affiliate links.

A Dotdash Meredith insider told Navarro that the company uses a tactic called "keyword swarming" to push high-quality independent sites off the top of Google and replace them with its own garbage reviews. When Dotdash Meredith finds an independent site that occupies the top results for a lucrative Google result, they "swarm a smaller site’s foothold on one or two articles by essentially publishing 10 articles [on the topic] and beefing up [Dotdash Meredith sites’] authority."

Dotdash Meredith has keyword swarmed a large number of topics. from air purifiers to slow cookers to posture correctors for back-pain:

https://housefresh.com/wp-content/uploads/2024/05/keyword-swarming-dotdash.jpg

The company isn't shy about this. Its own shareholder communications boast about it. What's more, it has competition.

Take Forbes, an actual news-site, which has a whole shadow-empire of web-pages reviewing products for puppies, dogs, kittens and cats, all of which link to high affiliate-fee-generating pet insurance products. These reviews are not good, but they are treasured by Google's algorithm, which views them as a part of Forbes's legitimate news-publishing operation and lets them draft on Forbes's authority.

This side-hustle for Forbes comes at a cost for the rest of us, though. The reviewers who actually put in the hard work to figure out which pet products are worth your money (and which ones are bad, defective or dangerous) are crowded off the front page of Google and eventually disappear, leaving behind nothing but semi-automated SEO garbage from Forbes:

https://twitter.com/ichbinGisele/status/1642481590524583936

There's a name for this: "site reputation abuse." That's when a site perverts its current – or past – practice of publishing high-quality materials to trick Google into giving the site a high ranking. Think of how Deadspin's private equity grifter owners turned it into a site full of casino affiliate spam:

https://www.404media.co/who-owns-deadspin-now-lineup-publishing/

The same thing happened to the venerable Money magazine:

https://moneygroup.pr/

Money is one of the many sites whose air purifier reviews Google gives preference to, despite the fact that they do no testing. According to Google, Money is also a reliable source of information on reprogramming your garage-door opener, buying a paint-sprayer, etc:

https://money.com/best-paint-sprayer/

All of this is made ten million times worse by AI, which can spray out superficially plausible botshit in superhuman quantities, letting spammers produce thousands of variations on their shitty reviews, flooding the zone with bullshit in classic Steve Bannon style:

https://escapecollective.com/commerce-content-is-breaking-product-reviews/

As Gizmodo, Sports Illustrated and USA Today have learned the hard way, AI can't write factual news pieces. But it can pump out bullshit written for the express purpose of drafting on the good work human journalists have done and tricking Google – the search engine 90% of us rely on – into upranking bullshit at the expense of high-quality information.

A variety of AI service bureaux have popped up to provide AI botshit as a service to news brands. While Navarro doesn't say so, I'm willing to bet that for news bosses, outsourcing your botshit scams to a third party is considered an excellent way of avoiding your journalists' wrath. The biggest botshit-as-a-service company is ASR Group (which also uses the alias Advon Commerce).

Advon claims that its botshit is, in fact, written by humans. But Advon's employees' Linkedin profiles tell a different story, boasting of their mastery of AI tools in the industrial-scale production of botshit:

https://housefresh.com/wp-content/uploads/2024/05/Advon-AI-LinkedIn.jpg

Now, none of this is particularly sophisticated. It doesn't take much discernment to spot when a site is engaged in "site reputation abuse." Presumably, the 12,000 googlers the company fired last year could have been employed to check the top review keyword results manually every couple of days and permaban any site caught cheating this way.

Instead, Google is has announced a change in policy: starting May 5, the company will downrank any site caught engaged in site reputation abuse. However, the company takes a very narrow view of site reputation abuse, limiting punishments to sites that employ third parties to generate or uprank their botshit. Companies that produce their botshit in-house are seemingly not covered by this policy.

As Navarro writes, some sites – like Forbes – have prepared for May 5 by blocking their botshit sections from Google's crawler. This can't be their permanent strategy, though – either they'll have to kill the section or bring it in-house to comply with Google's rules. Bringing things in house isn't that hard: US News and World Report is advertising for an SEO editor who will publish 70-80 posts per month, doubtless each one a masterpiece of high-quality, carefully researched material of great value to Google's users:

https://twitter.com/dannyashton/status/1777408051357585425

As Navarro points out, Google is palpably reluctant to target the largest, best-funded spammers. Its March 2024 update kicked many garbage AI sites out of the index – but only small bottom-feeders, not large, once-respected publications that have been colonized by private equity spam-farmers.

All of this comes at a price, and it's only incidentally paid by legitimate sites like Housefresh. The real price is borne by all of us, who are funneled by the 90%-market-share search engine into "review" sites that push low quality, high-price products. Housefresh's top budget air purifier costs $79. That's hundreds of dollars cheaper than the "budget" pick at other sites, who largely perform no original research.

Google search has a problem. AI botshit is dominating Google's search results, and it's not just in product reviews. Searches for infrastructure code samples are dominated by botshit code generated by Pulumi AI, whose chatbot hallucinates nonexistence AWS features:

https://www.theregister.com/2024/05/01/pulumi_ai_pollution_of_search/

This is hugely consequential: when these "hallucinations" slip through into production code, they create huge vulnerabilities for widespread malicious exploitation:

https://www.theregister.com/2024/03/28/ai_bots_hallucinate_software_packages/

We've put all our eggs in Google's basket, and Google's dropped the basket – but it doesn't matter because they can spend $20b/year bribing Apple to make sure no one ever tries a rival search engine on Ios or Safari:

https://finance.yahoo.com/news/google-payments-apple-reached-20-220947331.html

Google's response – laying off core developers, outsourcing to low-waged territories with weak labor protections and spending billions on stock buybacks – presents a picture of a company that is too big to care:

https://pluralistic.net/2024/04/04/teach-me-how-to-shruggie/#kagi

Google promised us a quid-pro-quo: let them be the single, authoritative portal ("organize the world’s information and make it universally accessible and useful"), and they will earn that spot by being the best search there is:

https://www.ft.com/content/b9eb3180-2a6e-41eb-91fe-2ab5942d4150

But – like the spammers at the top of its search result pages – Google didn't earn its spot at the center of our digital lives.

It cheated.

If you'd like an essay-formatted version of this post to read or share, here's a link to it on pluralistic.net, my surveillance-free, ad-free, tracker-free blog:

https://pluralistic.net/2024/05/03/keyword-swarming/#site-reputation-abuse

Image: freezelight (modified) https://commons.wikimedia.org/wiki/File:Spam_wall_-_Flickr_-_freezelight.jpg

CC BY-SA 2.0 https://creativecommons.org/licenses/by-sa/2.0/deed.en

[sayruq]

A new investigation by +972 Magazine and Local Call reveals that the Israeli army has developed an artificial intelligence-based program known as “Lavender,” unveiled here for the first time. According to six Israeli intelligence officers, who have all served in the army during the current war on the Gaza Strip and had first-hand involvement with the use of AI to generate targets for assassination, Lavender has played a central role in the unprecedented bombing of Palestinians, especially during the early stages of the war. In fact, according to the sources, its influence on the military’s operations was such that they essentially treated the outputs of the AI machine “as if it were a human decision.”

During the early stages of the war, the army gave sweeping approval for officers to adopt Lavender’s kill lists, with no requirement to thoroughly check why the machine made those choices or to examine the raw intelligence data on which they were based. One source stated that human personnel often served only as a “rubber stamp” for the machine’s decisions, adding that, normally, they would personally devote only about “20 seconds” to each target before authorizing a bombing — just to make sure the Lavender-marked target is male. This was despite knowing that the system makes what are regarded as “errors” in approximately 10 percent of cases, and is known to occasionally mark individuals who have merely a loose connection to militant groups, or no connection at all. Moreover, the Israeli army systematically attacked the targeted individuals while they were in their homes — usually at night while their whole families were present — rather than during the course of military activity. According to the sources, this was because, from what they regarded as an intelligence standpoint, it was easier to locate the individuals in their private houses. Additional automated systems, including one called “Where’s Daddy?” also revealed here for the first time, were used specifically to track the targeted individuals and carry out bombings when they had entered their family’s residences.

The Lavender machine joins another AI system, “The Gospel,” about which information was revealed in a previous investigation by +972 and Local Call in November 2023, as well as in the Israeli military’s own publications. A fundamental difference between the two systems is in the definition of the target: whereas The Gospel marks buildings and structures that the army claims militants operate from, Lavender marks people — and puts them on a kill list. In addition, according to the sources, when it came to targeting alleged junior militants marked by Lavender, the army preferred to only use unguided missiles, commonly known as “dumb” bombs (in contrast to “smart” precision bombs), which can destroy entire buildings on top of their occupants and cause significant casualties. “You don’t want to waste expensive bombs on unimportant people — it’s very expensive for the country and there’s a shortage [of those bombs],” said C., one of the intelligence officers. Another source said that they had personally authorized the bombing of “hundreds” of private homes of alleged junior operatives marked by Lavender, with many of these attacks killing civilians and entire families as “collateral damage.”

Remember, the Israeli occupation government considers all men over the age of 16 to be Hamas operatives hence why they've claimed to have killed over 9,000 of them (which matches the number of Palestinian men killed according to the Ministry of Health). So, when the article speaks of 'low level' or 'high level militants' they're likely speaking of civilians.

If Israel knew who Hamas fighters are, Oct 7th wouldn't have caught them off guard and they wouldn't still be fighting the Palestinian resistance every single day.