#AI algorithms

[starfruitcid3r]

Why is the Pinterest algorithm so bad?

I have few vices as far as social media goes; I limit myself to YouTube and Pinterest and those keep me sufficiently online. I go back to these sites again and again not because they are always exceedingly entertaining, but because I am able to see many of my interests compiled before my eyes for my investigation, especially on Pinterest. When I visit it, a colorful collage of Images is set before me. Whether I engage with these interests in the physical world or not is irrellevant; I am able to experience the dope rush associated with them vicariously through these images.

But today, I was browsing as normal and I refreshed the page to find myself looking at like some uncanny rendering of my tastes on paper. A chunk of black and white TikTok screenshots of trad goth folk, like six consecutive ads for car organizers on aliexpress, and multiple veins of Jenna Ortega stock photos. These things make sense to show me if you are an amnesiac robot. Indeed, my tastes include goth fashion, I was looking at car accessories, and Jenna Ortega is like cool I guess, but I haven't saved such overt trad-gothiness in a bit, I have not been overly interested in such elaborate car organization systems, and I saved a couple pins of Jenna Ortega's character from Beetlejuice Beetlejuice because I liked her costume weeks ago.

The Pinterest algorithm is known to me as addicting at best, temperamental and frustrating at worst, which is to say that sometimes it works rather beautifully, serving me my delicious images on a silver platter; things I want to see and have been looking at. But then, sometimes it turns on a dime and shows me beachy mainstream influencers I've never seen or heard of, softcore porn, and large clusters of similar images in waves that are inspired by one thing you saved ages ago. Even when the algorithm is working normally, search results are annoyingly hyper-curated (ex. if I've been looking at Tudor fashion and suddenly switch to something adjacent but distinct it will bleed into the search results terribly). Others have complained of looking at one thing one time and then getting only that recommended to them for a while, but I don't experience that too badly most of the time.

Another quite bothersome thing is the constant updates to the app. Sometimes bugs are fixed but the nigh constant unnecessary changes to the UI just introduces new ones and makes the user experience worse. Never have I once noticed an update and been like "Ah, what a nice change!" it was fine like 4 updates ago when we could swipe down on pins to go back to the previous page and scroll through super long images instead of just viewing them incomprehensibly as incredibly skinny rectangles scaled to fit entirely on a phone screen.

The ONE nice thing is it does seem that they have added a "AI modified" tag (seemingly desktop only) to some images, but this confuses me further because I have only just noticed it today and when I looked it up to find out more, I found a Reddit post from 4 months ago noting it. An artist commented that some of their original lineart had been tagged as "AI modified", and another commenter described how those using AI on the dl could easily circumvent this sad attempt by Pinterest to seem on top of things. My point here is a) Pinterest is bad as rolling out updates(?) and b) Pinterest is bad at moderating the AI content slowly overtaking the platform (but who isn't ig?).

Edit: images to exemplify my points

Jenna Ortega is extremely cool I'm sure but I saved 3 pins of her in the specific Beetlejuice II costume ages and ages ago

2. Who is she ?? ft. more Jenna Ortega and a comic I am not familiar with

3. I would get more images but it's back to normal now. WHY does the algorithm just periodically get horrible like a forest creature at a blood moon ??

[nothingxxtoxxsay]

"Eye in the sky"

[diversebeauties]

Kinda sucks that tumblr is censoring all your posts, like there is literally no need to do so. I hope tumblr fixes this issue soon

Whatever AI algorithm they're using lately is really bad. Terribly inaccurate. I think I do a pretty good job of curating a relatively safe feed on my own, I don't need it.

[thepastisalreadywritten]

Scholars of antiquity believe they are on the brink of a new era of understanding after researchers armed with artificial intelligence read the hidden text of a charred scroll that was buried when Mount Vesuvius erupted nearly 2,000 years ago.

Hundreds of papyrus scrolls held in the library of a luxury Roman villa in Herculaneum were burned to a crisp when the town was devastated by the intense blast of heat, ash and pumice that destroyed nearby Pompeii in AD79.

Excavations in the 18th century recovered more than 1,000 whole or partial scrolls from the mansion, thought to be owned by Julius Caesar’s father-in-law.

However, the black ink was unreadable on the carbonised papyri and the scrolls crumbled to pieces when researchers tried to open them.

The breakthrough in reading the ancient material came from the $1m Vesuvius Challenge, a contest launched in 2023 by Brent Seales, a computer scientist at the University of Kentucky, and Silicon Valley backers.

The competition offered prizes for extracting text from high-resolution CT scans of a scroll taken at Diamond, the UK’s national synchrotron facility in Oxfordshire.

On Monday, Nat Friedman, a US tech executive and founding sponsor of the challenge, announced that a team of three computer-savvy students, Youssef Nader in Germany, Luke Farritor in the US, and Julian Schilliger in Switzerland, had won the $700,000 (£554,000) grand prize after reading more than 2,000 Greek letters from the scroll.

Papyrologists who have studied the text recovered from the blackened scroll were stunned at the feat.

“This is a complete gamechanger,” said Robert Fowler, emeritus professor of Greek at Bristol University and chair of the Herculaneum Society.

“There are hundreds of these scrolls waiting to be read.”

Dr Federica Nicolardi, a papyrologist at the University of Naples Federico II, added:

“This is the start of a revolution in Herculaneum papyrology and in Greek philosophy in general. It is the only library to come to us from ancient Roman times.”

“We are moving into a new era,” said Seales, who led efforts to read the scrolls by virtually unwrapping the CT images and training AI algorithms to detect the presence of ink.

He now wants to build a portable CT scanner to image scrolls without moving them from their collections.

In October, Farritor won the challenge’s $40,000 “first letters” prize when he identified the ancient Greek word for “purple” in the scroll.

He teamed up with Nader in November, with Schilliger, who developed an algorithm to automatically unwrap CT images, joining them days before the contest deadline on 31 December.

Together, they read more than 2,000 letters of the scroll, giving scholars their first real insight into its contents.

“It’s been an incredibly rewarding journey,” said Youssef.

“The adrenaline rush is what kept us going. It was insane. It meant working 20-something hours a day. I didn’t know when one day ended and the next day started.”

“It probably is Philodemus,” Fowler said of the author.

“The style is very gnarly, typical of him, and the subject is up his alley.”

The scroll discusses sources of pleasure, touching on music and food – capers in particular – and whether the pleasure experienced from a combination of elements owes to the major or minor constituents, the abundant or the scare.

“In the case of food, we do not right away believe things that are scarce to be absolutely more pleasant than those which are abundant,” the author writes.

“I think he’s asking the question: what is the source of pleasure in a mix of things? Is it the dominant element, is it the scarce element, or is it the mix itself?” said Fowler.

The author ends with a parting shot against his philosophical adversaries for having “nothing to say about pleasure, either in general or particular."

Seales and his research team spent years developing algorithms to digitally unwrap the scrolls and detect the presence of ink from the changes it produced in the papyrus fibres.

He released the algorithms for contestants to build on in the challenge.

Friedman’s involvement proved valuable not only for attracting financial donors.

When Seales was meant to fly to the UK to have a scroll scanned, a storm blew in cancelling all commercial flights.

Worried they might lose their slot at the Diamond light source, Friedman hastily organised a private jet for the trip.

Beyond the hundreds of Herculaneum scrolls waiting to be read, many more may be buried at the villa, adding weight to arguments for fresh excavations.

"The same technology could be applied to papyrus wrapped around Egyptian mummies," Fowler said.

These could include everything from letters and property deeds to laundry lists and tax receipts, shining light on the lives of ordinary ancient Egyptians.

“There are crates of this stuff in the back rooms of museums,” Fowler said.

The challenge continues this year with the goal to read 85% of the scroll and lay the foundations for reading all of those already excavated.

Scientists need to fully automate the process of tracing the surface of the papyrus inside each scroll and improve ink detection on the most damaged parts.

“When we launched this less than a year ago, I honestly wasn’t sure it’d work,” said Friedman.

“You know, people say money can’t buy happiness, but they have no imagination. This has been pure joy. It’s magical what happened, it couldn’t have been scripted better."

Source: The Guardian

How the Herculaneum Papyri were carbonised in the Mount Vesuvius eruption – Video

5 February 2024

[ai-innova7ions]

Revolutionize Your Art with Leonardo AI!

Leonardo AI is revolutionizing artistic creation with its advanced algorithms that transform real-world ideas into stunning masterpieces. This dynamic platform empowers us to design imaginative game assets, including characters, artifacts, landscapes, conceptual visuals, and intricate architectures.

By merging cutting-edge technology with our creative fervor, Leonardo AI enables artists and designers to bring their visions to life. It injects depth and vibrancy into our projects, making it perfect for those looking to elevate their creative endeavors. Discover how this AI-driven toolset offers an unparalleled environment for artistic innovation!

#LeonardoAI #ArtisticInnovation

[takechargemedical]

The Rise of Generative AI in Medical Devices: Revolutionizing Healthcare Solutions

Generative AI is paving the way for groundbreaking advancements in the medical device industry. From accelerating design to enhancing patient outcomes, this transformative technology is reshaping the landscape of healthcare innovation. Here’s how generative AI is driving the future of medical device development:

1. Enhanced Device Design and Prototyping

Generative AI accelerates the ideation process, creating optimized device designs tailored to specific clinical needs. By analyzing vast datasets, AI identifies optimal configurations, reducing development time and costs so as to incorporate robust Gen-AI Driven Device Providers. 

2. Streamlined Manufacturing

AI-driven tools ensure precision in manufacturing, minimizing material waste and enhancing production efficiency. Generative models also predict potential flaws, ensuring robust device quality.

3. Improved Clinical Trial Outcomes

Medical devices integrated with AI capabilities support real-time patient monitoring, enabling adaptive clinical trials. Generative AI enhances patient recruitment and retention through predictive analytics, improving trial efficiency.

4. Personalized Patient Care

Generative AI enables devices to adapt to individual patient needs, such as adjusting insulin delivery in diabetes management. Personalized care leads to better patient compliance and improved outcomes.

5. Predictive Maintenance

AI algorithms forecast maintenance needs for complex medical equipment, minimizing downtime and ensuring continuous care delivery.

6. Regulatory Compliance Support

Generative AI aids in regulatory submissions by automating documentation and analyzing compliance data, expediting approval processes.

Embracing a Robust AI Future

Generative AI is not just a tool—it’s a transformative force driving innovation and efficiency in the medical device industry. Companies that harness its potential position themselves as leaders in delivering cutting-edge, patient-centric solutions.

Optimizing Healthcare for All For more insights on leveraging generative AI in medical devices, visit TakeChargeMedical.com and discover the future of healthcare innovation.

[techaiml]

Embracing the Revolution: Generative AI and Custom Software Development

Generative AI represents a paradigm shift in custom software development, offering unprecedented opportunities for innovation, efficiency, and creativity. By harnessing the power of AI algorithms, developers can accelerate the coding process, explore new design possibilities, and deliver highly customized solutions tailored to individual needs. As we embark on this journey of technological advancement, it's essential to embrace the transformative potential of generative AI while upholding ethical principles and ensuring inclusive and responsible development practices.

[techdriveplay]

The Impact of AI on Everyday Life: A New Normal

The impact of AI on everyday life has become a focal point for discussions among tech enthusiasts, policymakers, and the general public alike. This transformative force is reshaping the way we live, work, and interact with the world around us, making its influence felt across various domains of our daily existence. Revolutionizing Workplaces One of the most significant arenas where the impact…

[garudabluffs]

Kyle Chayka is a founding editor of Dirt. This essay is drawn from his new book Filterworld: How Algorithms Flattened Culture

Welcome to Filterworld

Kyle Chayka writes, “Our personal desires are constantly being anticipated, served in advance, and perhaps thus undercut.”

"it only matters that you’re still watching and clicking, building a map of your interests that make you more targetable. It’s long past time that we realized these systems are not benefitting culture. Creators and consumers alike are pushed into preset formats that we may ultimately have no organic interest in. By giving in to algorithmic feeds, we are letting tech companies determine our tastes. 

Flatness, like scalability, is efficient. The same culture flows through the same pipes to the same net-average consumer. But since when did efficiency become the sole metric by which we judge art? Filterworld represents the idea that the messiness of culture can be optimized and that only what is optimized for shareability is worthwhile. I disagree."

READ MORE https://dirt.fyi/article/2024/01/welcome-to-filterworld

The tyranny of the algorithm: why every coffee shop looks the same 16 Jan 2024

From the generic hipster cafe to the ‘Instagram wall’, the internet has pushed us towards a kind of global ubiquity – and this phenomenon is only going to intensify

"YouTube, founded in 2005, allowed anyone with a powerful enough internet connection to upload and share video clips. Instagram followed in 2010 and created a larger culture of sharing snapshots from newly mainstream iPhone cameras."

“Globalisation takes place only in capital and data,” the literary theorist Gayatri Chakravorty Spivak has written. “Everything else is damage control.” We talk about politics, culture and travel becoming globalised, but on a more fundamental level, Spivak is correct that what really flows across the planet are various forms of money and information: investments, corporations, infrastructure, server farms and the combined data of all the digital platforms, sluicing invisibly like wind or ocean currents between nations. We users voluntarily pumped our own information through this system, turning ourselves into flowing commodities, too.

This homogenisation is not just a phenomenon of our own moment; it is a consequence of changes that happened long before algorithmic social media feeds, and is just as likely to intensify in the future. After all, each time a grand flattening is announced the world somehow finds a way to get even flatter."

by Kyle Chayka READ MORE https://www.theguardian.com/news/2024/jan/16/the-tyranny-of-the-algorithm-why-every-coffee-shop-looks-the-same

How social media algorithms 'flatten' our culture by making decisions for us

JANUARY 17, 2024

LISTEN 36-Minute Listen READ MORE https://www.npr.org/2024/01/17/1224955473/social-media-algorithm-filterworld

[nothingxxtoxxsay]

"Heart of Gold"

The cross she bears is heavy, yet her heart, gleaming in gold, remains unyielding. With each breath she draws, she kindles a fragile ember of hope, daring to defy the relentless march of the encroaching shadows.

[mysmistree]

YouTube Is Terrible.

Sorry to be in whichever feed harping on about YouTube again but, It's kinda hit me that, aside from the video streaming function will I will confess is fairly top-notch, The website's kind of shit. And it's just been getting worse, and has no good competitor to prompt it to do better. (That I know of.) So there's the age verification thing rolling out in ten days that everyone knows about. There's also the fact they recently-ish rolled out shorts which, function okay enough but jump around a lot and load the wrong short if you pick out a set from a channel? And frequently will jump to the first or an arbitrary short in the list, and is so finnicky about moving about that it will sometimes not load the page altogether? There's also the ads, aggressive, frequent, and incredibly irritating. Constantly clogging up feeds and slowing down watch-times for things no one is interested in and gets skipped immediately. There's also the fact the search feature barely works and presumes more based on what it thinks you're looking for than what you're actually looking for, which is somehow more functional when you drunk-type your favorite video than when you spell the title out word for word. There's also the fact recommendations frequently over-corrects for misclicked videos or search terms, clogging ones feed with things they didn't want nor need. The fact channel pages are often missed, latest videos frequently flying completely under the radar even from channels one follows with great frequency.

The fact that several channels get completely mauled out of the picture because they aren't popular but support and raise rage-bait and divisive subjects even to people who avoid such like the plague, The fact it can be abused to hell by AI slop channels because it chose, a long time ago, to try and support more videos than it ever had any hope of moderating, And the fact its bot detection system took literal ages to be even half-decent because they couldn't be fucked to communicate or work with people on it for five minutes, just sort of shrugging and hoping it'd work itself out. I am not even a dependent creator. I don't rely on YouTube for my dime or dinner, but holy shit the stress that must bear is insane. Nevermind the bleeding age moderation because YouTube has long tried to make its ads some of the most valuable on the market but that winds up being all powerpoint buzzwords in the face of the fact it never fucking works properly and constantly winds up gutting videos about people just living their life because things about who they are landed on the political spectrum, which is fine if you want to police if kids see that (Or, with some others, mainstream it to them) but then demonetizing it because its sensitive on that matter? Are you TRYING to sell a five-year-old a shitty app that doesn't exist and is a scam? Just. It's so shit. I use it every single day because the dopamine hits are practically an addition in themselves and it's so fucking stupid. I want and intend to stop opening these tabs, and if that age gate goes through that'll be the easier occasion. But overall, I just wanted to vent about the fact that because its given us unbridled access to the worlds video making habits, YouTube has monopolized the watching market for so long that they've just let themselves fall apart.

[enterprise-cloud-services]

SEO teams understand intent, data, and language — making them the perfect leaders for companies adapting to the rise of AI-powered search engines.

[rubylogan15]

SEO teams understand intent, data, and language — making them the perfect leaders for companies adapting to the rise of AI-powered search engines.

[n-2-b]

In the current competitive digital landscape, it’s essential to go beyond simply knowing who your customers are; understanding their long-term value is critical for effective strategy development. This is where it plays a crucial role. By calculating CLV, businesses can gain insights into the future profitability of their customer relationships, allowing for more informed decision-making and targeted marketing efforts

In today’s hyper-competitive digital landscape, simply knowing who your customers are is no longer enough. The real strategy begins when you understand what each customer is worth over time, and that’s where Customer Lifetime Value comes into play.

What is CLV and Why Is It So Important?

Customer Lifetime Value refers to the total net profit a business expects to earn from a customer throughout the entire duration of their relationship. Unlike short-term metrics such as daily sales or web traffic, CLV shifts your focus to long-term business health. It helps brands move from chasing one-time buyers to cultivating loyal customers who return again and again.

Understanding CLV allows businesses to:

Allocate marketing budgets more wisely

Design customer experiences that foster loyalty

Increase profitability without increasing acquisition costs

To bring this to life, let’s explore 7 essential ways CLV can provide profitability in your digital marketing efforts.

[public-cloud-computing]

SEO teams understand intent, data, and language — making them the perfect leaders for companies adapting to the rise of AI-powered search engines.

[thesarkariyojna]

Gramhir.pro AI Image Generator: Creative Image Generation in 2025

Gramhir.pro AI Image Generator 2025: Unleash creative AI image generation. Craft high-quality visuals with this innovative generator tool. Explore AI image creation now! artificial intelligence is revolutionizing how we create visual content. The Gramhir Pro AI Photo Generator is a standout AI image generator, transforming simple text prompts into visually stunning images. Whether you’re a…