Feeds:
Posts
Comments

Posts Tagged ‘AI’

AI Reads the Unreadable

Posted in Uncategorized, tagged , , , , , , , , on November 24, 2025| Leave a Comment »

Photo: R.L. Easton, K. Knox, and W. Christens-Barry/Propietario del Palimpsesto de Arquímedes.
AI was used to translate this palimpsest with texts by Archimedes.

In general, I am wary of artificial intelligence, which one of its first developers has warned is dangerous. I use it to ask Google questions, but it’s a real nuisance in the English as a Second Language classes where I volunteer. Some students are tempted by the ease of using AI to do the homework, but of course, they learn nothing if they do that.

There’s another kind of translation, however, that AI seems good for: otherwise unreadable ancient texts.

Raúl Limón writes at El País, “In 1229, the priest Johannes Myronas found no better medium for writing his prayers than a 300-year-old parchment filled with Greek texts and formulations that meant nothing to him. At the time, any writing material was a luxury. He erased the content — which had been written by an anonymous scribe in present-day Istanbul — trimmed the pages, folded them in half and added them to other parchments to write down his prayers.

“In the year 2000, a team of more than 80 experts from the Walters Art Museum in Baltimore set out to decipher what was originally inscribed on this palimpsest — an ancient manuscript with traces of writing that have been erased. And, after five years of effort, they revealed a copy of Archimedes’ treatises, including The Method of Mechanical Theorems, which is fundamental to classical and modern mathematics.

“A Spanish study — now published in the peer-reviewed journal Mathematics — provides a formula for reading altered original manuscripts by using artificial intelligence. …

“Science hasn’t been the only other field to experience the effects of this practice. The Vatican Library houses a text by a Christian theologian who erased biblical fragments — which were more than 1,500-years-old — just to express his thoughts. Several Greek medical treatises have been deciphered behind the letters of a Byzantine liturgy. The list is extensive, but could be extended if the process of recovering these originals wasn’t so complex.

“According to the authors of the research published in Mathematics — José Luis Salmerón and Eva Fernández Palop — the primary texts within the palimpsests exhibit mechanical, chemical and optical alterations. These require sophisticated techniques — such as multispectral imaging, computational analysis, X-ray fluorescence and tomography — so that the original writing can be recovered. But even these expensive techniques yield partial and limited results. …

“The researchers’ model allows for the generation of synthetic data to accurately model key degradation processes and overcome the scarcity of information contained in the cultural object. It also yields better results than traditional models, based on multispectral images, while enabling research with conventional digital images.

“Salmerón — a professor of AI at CUNEF University in Madrid, a researcher at the Autonomous University of Chile and director of Stealth AI Startup — explains that this research arose from a proposal by Eva Fernández Palop, who was working on a thesis about palimpsests. At the time, the researcher was considering the possibility of applying new computational techniques to manuscripts of this sort.

“ ‘The advantage of our system is that we can control every aspect [of it], such as the level of degradation, colors, languages… and this allows us to generate a tailored database, with all the possibilities [considered],’ Salmerón explains.

“The team has worked with texts in Syriac, Caucasic, Albanian and Latin, achieving results that are superior to those produced by classical systems. The findings also include the development of the algorithm, so that it can be used by any researcher.

“This development isn’t limited to historical documents. ‘This dual-network framework is especially well-suited for tasks involving [cluttered], partially visible, or overlapping data patterns,’ the researcher clarifies. These conditions are found in medical imaging, remote sensing, biological microscopy and industrial inspection systems, as well as in the forensic investigation of images and documents. …

“The researchers themselves admit that there are limitations to their proposed method for examining palimpsests: ‘The approach shows degraded performance when processing extremely faded texts with contrast levels below 5%, where essential stroke information becomes indistinguishable from crumbling parchment. Additionally, the model’s effectiveness depends on careful script balancing during the training phase, as unequal representation of writing systems can make the deep-learning features biased toward more frequent scripts.’ ”

More at El País, here. What is your view of AI? All good? Dangerous? OK sometimes? I can’t stop thinking about the warning from Geoffrey Hinton, the ‘godfather of AI,’ that it could wipe out humanity altogether. 

Read Full Post »

Learning to Spot AI Text

Posted in Uncategorized, tagged , , , , , on September 22, 2025| 7 Comments »

Today is my second online ESL (English as a Second Language) class for the ’25-’26 school year. I assist a more experienced teacher once a week — have been doing so for nearly ten years. One task she likes me to do is to go over the writing homework that students put on an edublog.

Lately, it feels like these otherwise highly motivated adults may not be learning much about writing English. Often they seem to have copied from Google Translate or another AI program. What I want to see is a few mistakes in their answers. At the same time, I am wary of accusing anyone of not doing their own work.

Today’s article didn’t give me a clear answer to my ESL situation, but I was intrigued to learn about programs that help identify who the real writer of a book was or whether AI was used in a journal article.

Roger J. Kreuz, associate dean and professor of psychology, University of Memphis, writes at the Conversation that although it’s common to use chatbots “to write computer code, summarize articles and books, or solicit advice … chatbots are also employed to quickly generate text from scratch, with some users passing off the words as their own.

“This has, not surprisingly, created headaches for teachers tasked with evaluating their students’ written work. It’s also created issues for people seeking advice on forums like Reddit, or consulting product reviews before making a purchase.

“Over the past few years, researchers have been exploring whether it’s even possible to distinguish human writing from artificial intelligence-generated text. … Research participants recruited for a 2021 online study, for example, were unable to distinguish between human- and ChatGPT-generated stories, news articles and recipes.

“Language experts fare no better. In a 2023 study, editorial board members for top linguistics journals were unable to determine which article abstracts had been written by humans and which were generated by ChatGPT. And a 2024 study found that 94% of undergraduate exams written by ChatGPT went undetected by graders at a British university. …

“A commonly held belief is that rare or unusual words can serve as ‘tells’ regarding authorship, just as a poker player might somehow give away that they hold a winning hand.

“Researchers have, in fact, documented a dramatic increase in relatively uncommon words, such as ‘delves’ or ‘crucial,’ in articles published in scientific journals over the past couple of years. This suggests that unusual terms could serve as tells that generative AI has been used. It also implies that some researchers are actively using bots to write or edit parts of their submissions to academic journals. …

“In another study, researchers asked people about characteristics they associate with chatbot-generated text. Many participants pointed to the excessive use of em dashes – an elongated dash used to set off text or serve as a break in thought – as one marker of computer-generated output. But even in this study, the participants’ rate of AI detection was only marginally better than chance.

“Given such poor performance, why do so many people believe that em dashes are a clear tell for chatbots? Perhaps it’s because this form of punctuation is primarily employed by experienced writers. In other words, people may believe that writing that is ‘too good’ must be artificially generated.

“But if people can’t intuitively tell the difference, perhaps there are other methods for determining human versus artificial authorship.

“Some answers may be found in the field of stylometry, in which researchers employ statistical methods to detect variations in the writing styles of authors.

“I’m a cognitive scientist who authored a book on the history of stylometric techniques. In it, I document how researchers developed methods to establish authorship in contested cases, or to determine who may have written anonymous texts.

“One tool for determining authorship was proposed by the Australian scholar John Burrows. He developed Burrows’ Delta, a computerized technique that examines the relative frequency of common words, as opposed to rare ones, that appear in different texts.

“It may seem counterintuitive to think that someone’s use of words like ‘the,’ ‘and’ or ‘to’ can determine authorship, but the technique has been impressively effective.

“Burrows’ Delta, for example, was used to establish that Ruth Plumly Thompson, L. Frank Baum’s successor, was the author of a disputed book in the Wizard of Oz series. It was also used to determine that love letters attributed to Confederate Gen. George Pickett were actually the inventions of his widow, LaSalle Corbell Pickett.

“A major drawback of Burrows’ Delta and similar techniques is that they require a fairly large amount of text to reliably distinguish between authors. A 2016 study found that at least 1,000 words from each author may be required. A relatively short student essay, therefore, wouldn’t provide enough input for a statistical technique to work its attribution magic.

“More recent work has made use of what are known as BERT language models, which are trained on large amounts of human- and chatbot-generated text. The models learn the patterns that are common in each type of writing, and they can be much more discriminating than people: The best ones are between 80% and 98% accurate.

“However, these machine-learning models are ‘black boxes’ – that is, we don’t really know which features of texts are responsible for their impressive abilities. Researchers are actively trying to find ways to make sense of them, but for now, it isn’t clear whether the models are detecting specific, reliable signals that humans can look for on their own.

“Another challenge for identifying bot-generated text is that the models themselves are constantly changing – sometimes in major ways.

“Early in 2025, for example, users began to express concerns that ChatGPT had become overly obsequious, with mundane queries deemed ‘amazing’ or ‘fantastic.’ OpenAI addressed the issue by rolling back some changes it had made.

“Of course, the writing style of a human author may change over time as well, but it typically does so more gradually.

“At some point, I wondered what the bots had to say for themselves. I asked ChatGPT-4o: ‘How can I tell if some prose was generated by ChatGPT? Does it have any “tells,” such as characteristic word choice or punctuation?’

“[It provided] me with a 10-item list, replete with examples. These included the use of hedges – words like ‘often’ and ‘generally’ – as well as redundancy, an overreliance on lists and a ‘polished, neutral tone.’ It did mention ‘predictable vocabulary,’ which included certain adjectives such as ‘significant’ and ‘notable,’ along with academic terms like ‘implication’ and ‘complexity.’ However, though it noted that these features of chatbot-generated text are common, it concluded that ‘none are definitive on their own.’ ” More at the Conversation, here.

If I were in the room with students, I could more or less stand over them and see how they go about writing. But these are adults, after all, and they want to learn, so the goal is to persuade them how learning is more likely to happen. Let me know if you have ideas that could help me.

Read Full Post »

Authors Fight Back Against AI

Posted in Uncategorized, tagged , , , , , , on August 23, 2025| 1 Comment »

Photo: Steve Johnson.
Real books start with a human, a human with feelings.

Blogger Asakiyume is an activist against AI. And no wonder. She’s an author, and an especially creative one. Believe me, what her brain comes up with, no one else’s brain ever could! AI, however, just copies what has come before.

So right now, as other published authors are uniting against AI robot writers, she’s in good company.

Chloe Veltman reports at National Public Radio, “A group of more than 70 authors including Dennis Lehane, Gregory Maguire and Lauren Groff released an open letter on Friday about the use of AI on the literary website Lit Hub. It asked publishing houses to promise ‘they will never release books that were created by machines.’

“Addressed to the ‘big five’ U.S. publishers — Penguin, Random House, HarperCollins, Simon & Schuster, Hachette Book Group, and Macmillan — as well as ‘other publishers of America,’ the letter elicited more than 1,100 signatures on its accompanying petition in less than 24 hours. Among the well-known signatories after the letter’s release are Jodi Picoult, Olivie Blake and Paul Tremblay.

“The letter contains a list of direct requests to publishers concerning a wide array of ways in which AI may already — or could soon be — used in publishing. It asks them to refrain from publishing books written using AI tools built on copyrighted content without authors’ consent or compensation, to refrain from replacing publishing house employees wholly or partially with AI tools, and to only hire human audiobook narrators — among other requests. …

“The letter states, ‘AI is an enormously powerful tool, here to stay, with the capacity for real societal benefits — but the replacement of art and artists isn’t one of them.’

“Until now, authors have mostly expressed their displeasure with AI’s negative impacts on their work by launching lawsuits against AI companies rather than addressing publishing houses directly. Ta-Nehisi Coates, Michael Chabon, Junot Díaz and the comedian Sarah Silverman are among the biggest names involved in ongoing copyright infringement cases against AI players.

“Some of these cases are already starting to render rulings: Earlier this week, federal judges presiding over two such cases ruled in favor of defendants Anthropic AI and Meta, potentially giving AI companies the legal right under the fair use doctrine to train their large language models on copyrighted works — as long as they obtain copies of those works legally.

“Young adult fiction author Rioghnach Robinson, who goes by the pen name Riley Redgate … said, ‘Without publishers pledging not to generate internally competitive titles, nothing’s stopping publishing houses from AI-generating their authors out of existence. We’re hopeful that publishers will act to protect authors and industry workers from, specifically, the competitive and labor-related threats of AI.’

“The authors said the ‘existential threat’ of AI isn’t just about copyright infringement. Copycat books that appear to have been written by AI and are attached to real authors who didn’t write them have proliferated on Amazon and other platforms in recent years.

“The rise of AI audio production within publishing is another big threat addressed in the letter. Many authors make extra money narrating their own books. And the rise of machine narration and translation is an even greater concern for human voice actors and translators. For example, major audio books publisher Audible recently announced a partnership with publishers to expand AI narration and translation offerings. …

“Audible CEO Bob Carrigan said as part of the announcement, ‘We’ll be able to bring more stories to life — helping creators reach new audiences while ensuring listeners worldwide can access extraordinary books that might otherwise never reach their ears.’

“Robinson acknowledged the steps publishers have taken to help protect writers.

” ‘Many individual contracts now have AI opt-out clauses in an attempt to keep books out of AI training datasets, which is great,’ Robinson noted. But she said publishers should be doing much more to defend their writers against the onslaught of AI.”

More at NPR, here.

Read Full Post »

Translators of Poetry: Line Up!

Posted in Uncategorized, tagged , , , , , on August 12, 2025| 3 Comments »

Photo: Fitzcarraldo Editions.
Three publishing companies have launched the biennial Poetry in Translation prize, which will award an advance of $5,000 to be shared equally between poet and translator.

Anyone who has used Google Translate for a simple sentence knows that AI is not going to be doing quality translations of whole books anytime soon. There is too much subtlety needed.

And if that’s true for, say, a murder mystery, imagine how important a human translator is for poetry!

That’s why a new prize for poetry translation from publishers in the UK, Australia, and the US is arriving just in time — before the world gets lulled into thinking an AI translation is just fine.

Ella Creamer reports at the Guardian, “A new poetry prize for collections translated into English is opening for entries. …

“Publishers Fitzcarraldo Editions, Giramondo Publishing and New Directions have launched the biennial Poetry in Translation prize, which will award an advance of $5,000 (£3,700) to be shared equally between poet and translator.

“The winning collection will be published in the UK and Ireland by Fitzcarraldo Editions, in Australia and New Zealand by Giramondo and in North America by New Directions.

“ ‘We wanted to open our doors to new poetry in translation to give space and gain exposure to poetries we may not be aware of,’ said Fitzcarraldo poetry editor Rachael Allen. …

“The prize announcement comes amid a sales boom in translated fiction in the UK. Joely Day, Allen’s co-editor at Fitzcarraldo, believes that ‘the space the work of translators has opened up in the reading lives of English speakers through the success of fiction in translation will also extend to poetry.’ …

“Fitzcarraldo has published translated works by Nobel prize winners Olga Tokarczuk, Jon Fosse and Annie Ernaux. ‘Our prose lists have always maintained a roughly equitable balance between English-language and translation, and some of our greatest successes have been books in translation,’ said Day. ‘We’d like to bring the same diversity of voices to our poetry publishing.’ …

“The prize is open to living poets from around the world, writing in any language other than English.

“The prize is being launched to find works ‘which are formally innovative, which feel new, which have a strong and distinctive voice, which surprise and energize and move us,’ said Day. ‘My personal hope is that the prize reaches fledgling or aspiring translators and provides an opening for them.’ …

“Submissions will be open from 15 July to 15 August. A shortlist will be announced later this year, with the winner announced in January 2026 and publication of the winning collection scheduled for 2027.

“The ‘unique’ award ‘brings poetry from around the world into English, and foregrounds the essential role of translation in our literature,’ said Nick Tapper, associate publisher at Giramondo. ‘Its global outlook will bring new readers to poets whose work deserves wide and sustained attention.’ ”

More at the Guardian, here. I hope a certain blogger who translates Vietnamese poetry into English will apply for that prize.

Read Full Post »

How AI Is Helping a Small Publication

Posted in Uncategorized, tagged , , , , , , , , on August 7, 2025| 5 Comments »

Photo: MinnPost.
Cynthia Tu of Sahan Journal is using Chat GPT to improve revenue streams.

A few times in the past, I’ve had reason to link to a story at Sahan Journal, a nonprofit newsroom serving immigrants and communities of color in Minnesota. Now NiemanLab, a website about journalism, links to an article on a surprising development at the small publisher.

Lev Gringauz, reporting at MinnPost via NiemanLab, writes “As journalists around the world experiment with artificial intelligence, many newsrooms have common, often audience-facing, ideas for what to try.

“They range from letting readers talk to chatbots trained on reporting, to turning written stories into audio, creating story summaries and, infamously, generating entire articles using AI — a use case vehemently rejected by many journalists.

“But Sahan Journal, the nonprofit newsroom serving immigrants and communities of color in Minnesota, wanted to try something different.

“ ‘We’re less enthusiastic, more skeptical, about using AI to generate editorial content,’ said Cynthia Tu, Sahan Journal’s data journalist and AI specialist.

“Instead, the outlet has been working on ways to support internal workflows with AI. Now, it’s even testing a custom ChatGPT bot to help pitch Sahan Journal to prospective advertisers and sponsors. …

“While AI has plenty of ethical and technical issues, Tu’s work highlights another important aspect: The intended users — in this case, the Sahan Journal team.

“ ‘A lot of … this experiment is less of a technical challenge,’ Tu said. ‘It’s more like, how do you make [AI] fit in the human system more flawlessly? And how do you train the human to use this tool in a way that it was intended?’

“Sahan Journal’s AI experimentation, and Tu’s job, are supported by a partnership between the American Journalism Project, a national nonprofit helping local newsrooms, and ChatGPT creator OpenAI. …

“Liam Andrew, technology lead for the AJP’s Project & AI Studio, sees part of his job as helping newsrooms overcome hesitancy around AI. …

“Tu joined Sahan Journal fresh from a Columbia Journalism School master’s program in data journalism. She had played a little with chatbots, but otherwise didn’t have much experience working with AI. …

“For one investigation, Tu used a Google AI tool to process the financial data of charter schools in Minnesota. Thinking about how to save time on backend workflows, Tu then helped Sahan Journal generate story summaries, tailored for Instagram carousels, with ChatGPT. …

“ ‘You need to know what the workflow of the organization looks like…[and how] you push for change within a department when they’ve already been doing [something] for the past five years using a manual or human labor way.’

“That knowledge came in handy when finally tackling Tu’s core AI project: improving Sahan Journal’s revenue.

“The project stemmed from an anonymized database of audience insights, which included demographic information and interests. While an important resource, Sahan Journal’s small revenue team didn’t have the time to figure out how to leverage it. …

” ‘What if AI could feed two birds with one scone? A custom ChatGPT bot could process the audience data and personalize a media kit for clients. But it needed to work without being an extra burden on the revenue staff. …

“The magic of AI chatbots like ChatGPT is that you don’t need to know how to code to use them. Just type in a prompt and get rolling. …

“Less magically, AI chatbots can be hard to keep in line for specific tasks. Designed to be eager helpers, they hallucinate false results and stubbornly twist instructions in an attempt to please.

“Troubleshooting those issues was no simple task for Tu.

“The custom revenue chatbot struggled to keep Tu’s preferred formatting, and hallucinated audience data. The bot would also intermix results from the internet that Tu had not asked for. None of that was ideal for a tool that should work reliably for the revenue team.

“ ‘I was kind of jumping through hoops and telling it multiple times, “Please do not reference anything else on the internet,” ‘ Tu said. …

“Working with chatbots is an exercise in prompt engineering — mostly a trial-and-error process of figuring out what specific instructions will get the preferred result. As Tu said, ‘lazy questions lead to lazy answers.’ … Eventually, Tu settled on a reliable set of prompts.

“The custom chatbot takes about 20 seconds to find relevant data from the audience database — for example, pulling up how much of Sahan Journal’s audience cares about public transportation. Then it creates a summary for a media kit tailored to potential clients.

“The chatbot also double-checks its work by referencing the database again, making sure its output matches reality. And part of the database is shown for users to manually see the chatbot isn’t hallucinating. …

“Earlier this year, Tu introduced the final version of the revenue bot to Sahan Journal’s team. …

“By mid-April, the Sahan Journal revenue team had used the custom chatbot on six sales pitches, with three successfully leading to ads placed on the site. …

“But there’s a larger question hanging over this work: Is it sustainable? In a way, newsroom experiments with AI exist in a bubble.

“ ‘Everything is kind of tied to a grant,’ Tu said, referencing the AJP-OpenAI partnership that supports her work. But grants come and go as donor interests (and financials) change.”

The other unknowns are weighed at NiemanLabs, here.

Read Full Post »

Duolingo Adds isiZulu. And More.

Posted in Uncategorized, tagged , , , , , on June 14, 2025| 11 Comments »

The Duolingo bird can be very encouraging to a language learner. But it can also get angry.

With Suzanne’s family leaving soon for six months in Stockholm, I’ve been trying to learn some Swedish. I hope to try it on my grandchildren come next January. So it’s daily Duolingo for me. If I ever get to the point where I can understand Erik when he uses Swedish with the kids, I might also try expanding my French. I like the way the silly Duolingo bird cheers me on.

I was surprised to learn how many new languages the app has been adding lately. In the beginning, it didn’t even have Swedish. Now, according to an article in the Verge, it’s adding things like Maori, Tagalog, Haitian Creole, and isiZulu.

Jay Peters reports, “Duolingo is ‘more than doubling’ the number of courses it has available, a feat it says was only possible because it used generative AI to help create them in ‘less than a year.’

“The company [said] that it’s launching 148 new language courses. ‘This launch makes Duolingo’s seven most popular non-English languages – Spanish, French, German, Italian, Japanese, Korean, and Mandarin – available to all 28 supported user interface (UI) languages,’ dramatically expanding learning options for over a billion potential learners worldwide’ … the company writes.

“Duolingo says that building one new course historically has taken ‘years,’ but the company was able to build this new suite of courses more quickly ‘through advances in generative AI, shared content systems, and internal tooling.’ The new approach is internally called ‘shared content,’ and the company says it allows employees to make a base course and quickly customize it. …

“ ‘Now, by using generative AI to create and validate content, we’re able to focus our expertise where it’s most impactful, ensuring every course meets Duolingo’s rigorous quality standards,’ Duolingo’s senior director of learning design, Jessie Becker, says in a statement.

“The announcement follows a recent memo sent by cofounder and CEO Luis von Ahn to staff saying that … it would ‘gradually stop using contractors to do work that AI can handle.’ AI use will now be evaluated during the hiring process and as part of performance reviews, and von Ahn says that ‘headcount will only be given if a team cannot automate more of their work.’

“Spokesperson Sam Dalsimer tells The Verge in response to questions sent following von Ahn’s memo. ‘We’ve already been moving in this direction, and it has been game-changing for our company. One of the best decisions we made recently was replacing a slow, manual content creation process with one powered by AI, under the direction of our learning design experts. That shift allowed us to create and launch 148 new language courses today.’ …

“Dalsimer acknowledges that there have been ‘negative reactions’ to von Ahn’s memo. Dalsimer also notes that Duolingo has ‘no intention to reduce full-time headcount or hiring’ and that ‘any changes to contractor staffing will be considered on a case-by-case basis.’ “

Hmm. That is giving me pause. But I do like the app and the way that for English-speaking students like me, Duolingo starts out with some vocabulary that sounds like English. It makes me wonder if it does the same for learners who come from other languages. That could be really tricky.

Have you used Duolingo? I know that blogger Asakiyume, a mega language learner, used Duolingo to add Spanish and Portuguese to what she already knew in Japanese and more obscure languages. One thing I know for sure: she won’t like that Duolingo contractors will lose jobs thanks to AI.

More at the Verge, here.

Read Full Post »

AI De-Ages Tom Hanks

Posted in Uncategorized, tagged , , , , , , , , , on January 29, 2025| 3 Comments »

Photo: Everett Collection.
A de-aged version of actors Tom Hanks and Robin Wright, created by artificial intelligence for the 2024 film Here.

We are well into the age of AI, and I certainly hope that doesn’t mean we’re going to realize the dire warnings of one of its pioneers but just use it in relatively harmless ways.

Today’s story is about using AI to “de-age” actors in a movie covering 60 years.

Benj Edwards writes at Wired, “Here, a $50 million Robert Zemeckis–directed film [used] real-time generative AI face transformation techniques to portray actors Tom Hanks and Robin Wright across a 60-year span, marking one of Hollywood’s first full-length features built around AI-powered visual effects.

“The film adapts a 2014 graphic novel set primarily in a New Jersey living room across multiple time periods. Rather than cast different actors for various ages, the production used AI to modify Hanks’s and Wright’s appearances throughout.

“The de-aging technology comes from Metaphysic, a visual effects company that creates real time face swapping and aging effects. During filming, the crew watched two monitors simultaneously: one showing the actors’ actual appearances and another displaying them at whatever age the scene required.

“Metaphysic developed the facial modification system by training custom machine-learning models on frames of Hanks’ and Wright’s previous films. This included a large dataset of facial movements, skin textures, and appearances under varied lighting conditions and camera angles. …

“Unlike previous aging effects that relied on frame-by-frame manipulation, Metaphysic’s approach generates transformations instantly by analyzing facial landmarks and mapping them to trained age variations. … Traditional visual effects for this level of face modification would reportedly require hundreds of artists and a substantially larger budget closer to standard Marvel movie costs.

“This isn’t the first film that has used AI techniques to de-age actors. ILM’s approach to de-aging Harrison Ford in 2023’s Indiana Jones and the Dial of Destiny used a proprietary system called Flux with infrared cameras to capture facial data during filming, then old images of Ford to de-age him in postproduction. By contrast, Metaphysic’s AI models process transformations without additional hardware and show results during filming. …

“Meanwhile, as we saw with the SAG-AFTRA union strike [in 2023], Hollywood studios and unions continue to hotly debate AI’s role in filmmaking. While the Screen Actors Guild and Writers Guild secured some AI limitations in recent contracts, many industry veterans see the technology as inevitable. …

“Even so, the New York Times says that Metaphysic’s technology has already found use in two other 2024 releases. Furiosa: A Mad Max Saga employed it to re-create deceased actor Richard Carter’s character, while Alien: Romulus brought back Ian Holm’s android character from the 1979 original. Both implementations required estate approval under new California legislation governing AI recreations of performers, often called deepfakes. …

“Robert Downey Jr. recently said in an interview that he would instruct his estate to sue anyone attempting to digitally bring him back from the dead for another film appearance. But even with controversies, Hollywood still seems to find a way to make death-defying (and age-defying) visual feats take place onscreen — especially if there is enough money involved.”

What could go wrong?

The first thing I think of is fewer job opportunities for actors who play younger versions of stars. Still, I’d love to see an AI child version of the actress who plays Astrid in the French crime show of the same name, because I think it would look more natural than the mimicking girl they’ve got. (Awesome tv, by the way. Check it out on PBS Passport.)

More at Wired, here. This story originally appeared on Ars Technica.

Read Full Post »

Fish Highway in Stockholm

Posted in Uncategorized, tagged , , , , , on January 14, 2025| Leave a Comment »

Photo: TiVa.
Installation of the fish counter at Gamla Stan in Slussen. The new fish highway in Stockholm has some of the first fish passages between the Baltic Sea and Lake Mälaren at Söderström in nearly 400 years.

Today’s story is about how Sweden is giving a helping hand to migrating fish that are not strong swimmers.

TiVa, an AI-powered fish-counting company, reports, “In mid-2024, a TiVA FC was installed in connection to the newly built fish migration path at Slussen in Stockholm. This long-awaited passage allows fish to freely migrate between Lake Mälaren and Saltsjön at Söderström for the first time in almost 400 years. As part of the reconstruction of Slussen, a new fish migration path has been constructed under the northern sluice quay on the side of Gamla Stan. The old Nils Ericson sluice has been converted into a passage to facilitate the free movement of weak-swimming species such as eel, roach, and perch – species that were previously hindered by human infrastructure.

“The TiVA FC fish counter delivers [improved] results, both in image quality and AI-based species and length classification. … The TiVA FC at Slussen is connected to our cloud platform fiskdata.se. Here, data is available for both the client and, in this case, for the public. … The City of Stockholm has chosen to broadcast a live stream via TiVA’s YouTube channel. Shorter streams can also be broadcasted to other platforms, like Facebook, depending on the client’s needs.

“The City of Stockholm will install an informational screen for ‘Fish TV’ on the crane structure by the fish counter. Passersby in Gamla Stan will be able to learn more about the project, see selected videos, and get updates on the latest migrations.” More at the TiVa website, here.

And from Stockholm’s website: “You can watch online the fish swimming between Lake Mälaren and Saltsjön [at fiskdata.se].

“Moving between different areas is a natural part of life for many fish species. They migrate from their breeding grounds to spawning grounds to reproduce. When humans have blocked various watercourses, this has prevented fish species from passing through. To protect the fish and promote the environment, watercourses can be restored, or, as here at Slussen, a fish migration route can be opened up.

“The primary purpose of the fish migration route is to enable passage between Lake Mälaren and Saltsjön for [fish] that do not jump, which is basically all species except salmon and sea trout. By building this route, we hope that species such as eel, roach and perch will once again be able to pass here.

“The trail is designed by experts to mimic as natural an environment as possible. Stones of various sizes have been carefully placed along the trail. Some of the stones come from Gustav Vasa’s 16th-century defensive wall, which was found during the excavations of Södermalmstorg in 2022. The water flow needs to be calm so that the fish can stop and rest. There is lighting here so that the fish can swim in pleasant light.

“The fish walking trail is located under the quay on the Old Town side and is not visible from the outside. But you can watch the fish swimming through at fiskdata.se.” More here.

Looking for comments — from Swedes and fish lovers everywhere.

Read Full Post »

AI Finds Unknown Glyphs

Posted in Uncategorized, tagged , , , , , , on December 12, 2024| 2 Comments »

Photo: Instituto Universitario Yamagata de Nazca.
Some of the new geoglyphs found in Nazca. With their lines eroded by the passage of time, AI has achieved in months what used to take decades.

Let’s have kind word for scary old artificial intelligence and how it has, for example, helped to uncover 303 new geoglyphs in the Nazca desert. (By which I don’t mean to say AI doesn’t have serious potential dangers.)

In an El País archaeological article from Peru, Miguel Ángel Criado reports, “With the help of an artificial intelligence (AI) system, a group of archaeologists has uncovered in just a few months almost as many geoglyphs in the Nazca Desert (Peru) as those found in all of the last century. The large number of new figures has allowed the researchers to differentiate between two main types, and to offer an explanation of the possible reasons or functions that led their creators to draw them on the ground more than 2,000 years ago.

“The Nazca desert, with an area of about 1,900 square miles and an average altitude of 500 meters above sea level, has very special climatic conditions. It hardly ever rains, the hot air blocks the wind and the dry land has prevented the development of agriculture or livestock. Combined, all this has allowed a series of lines and figures, formed by stacking and aligning pebbles and stones, to be preserved for centuries. 

“The first layer of soil is made up of a blanket of small reddish stones that, when lifted, reveal a second yellowish layer. This difference in color is the basis of the geoglyphs and is what was used to create them by the ancient Nazca civilization. Some are straight lines stretching several miles. Others are geometric shapes or rectilinear figures, also huge in size.

“The other major category includes the so-called relief-type geoglyphs, which are smaller. In the 1930s, Peruvian aviators discovered the first ones, and by the end of the century more than a hundred had been identified, such as the hummingbird, the frog and the whale. Since 2004, supported by high-resolution satellite images, Japanese archaeologists have discovered 318 more, almost all of them high-profile geoglyphs. The same team, led by Masato Sakai, a scientist from Yamagata University (Japan), has discovered 303 new geoglyphs in a single campaign, supported by artificial intelligence. …

“ ‘The Nazca Pampa is a vast area covering more than 400 square kilometres and no exhaustive study has been carried out,’ the Japanese scientist recalls. Only the northern part, where the large linear geoglyphs are concentrated, ‘has been studied relatively intensively.’ … But scattered throughout the rest of the desert are many relief-type figures that are smaller and that the passage of time has made more difficult to detect.

“Convinced that there were many more, Sakai and his team contacted IBM’s artificial intelligence division. … They had high-resolution images obtained from airplanes or satellites of all of Nazca, but with a resolution of up to a few centimeters per pixel, the human eye would have needed years, if not decades, to analyze all the data. They left that job to the AI system. Although it was not easy to train its artificial vision … with so few previous images and so different from each other, the machine proposed 1,309 candidates. The figure came from a previous selection also made by the AI with 36 images for each candidate. With this selection, the researchers carried out a field expedition between September 2022 and February 2023. The result, as reported in the scientific journal PNAS, is 303 new geoglyphs added to this cultural heritage of humanity. All are relief-type geoglyphs.

“The newly discovered shapes bring the total number found in Nazca to 50 line-type and 683 relief-type geoglyphs, some geometric and others forming figures. The large amount has allowed the authors of this work to detect patterns and differences. Almost all of the former (the monkey, the condor, the cactus…) represent wild animals or plants. However, among the latter, almost 82% show human elements or elements modified by humans. ‘[There] are scenes of human sacrifice,’ says Sakai. …

“The accumulation of data that has made this work possible brings to light a double connection. On the one hand, these relief-type forms are found a few meters from one of the many paths that cross the desert … paths created by the passage of people until a path is created. According to the authors of the study, these creations were made to be seen by travelers.

“On the other hand, the large linear figures appear very close, also meters away, from one of the many straight lines that cut through the pampas. Here, according to Sakai, the symbolic value rules: ‘The line-type geoglyphs are drawn at the start and end points of the pilgrimage route to the Cahuachi ceremonial center. They were ceremonial spaces with shapes of animals and other figures. Meanwhile, the relief-type geoglyphs can be observed when walking along the paths.’

“Cahuachi was the seat of spiritual power of the Nazca culture between from around 100 BC to 500 AD and, for the authors, the large forms could be ceremonial stops on the pilgrimage to or from there.

“These explanations do not necessarily rule out, according to the authors, other possible functions that have been attributed to the Nazca lines and figures, such as being calendars, astronomical maps or even systems for capturing the little water that fell.”

Things do get fuzzy when we start to interpret ancient signs. Read more at El Pais, here. No firewall.

Read Full Post »

Popular AI Words to Watch For

Posted in Uncategorized, tagged , , , , on September 12, 2024| 12 Comments »

Image: CNN.
When you see phrases like “intricate interplay,” a concept I illustrate with this CNN image, suspect AI. There are lots of other common usages to be alert to.

Hello, Word People. That is, People Who Are Fascinated by Words. When you do online searches in the future, will you know what words and phrases may indicate AI behind the scenes? Kyle Orlando has some answers at Ars Technica.

“Thus far, even AI companies have had trouble coming up with tools that can reliably detect when a piece of writing was generated using a large language model [LLM]. Now, a group of researchers has established a novel method for estimating LLM usage across a large set of scientific writing by measuring which ‘excess words’ started showing up much more frequently during the LLM era (i.e., 2023 and 2024). The results ‘suggest that at least 10 percent of 2024 abstracts were processed with LLMs,’ according to the researchers.

“In a preprint paper posted earlier this month, four researchers from Germany’s University of Tübingen and Northwestern University said they were inspired by studies that measured the impact of the Covid-19 pandemic by looking at excess deaths compared to the recent past. By taking a similar look at ‘excess word usage’ after LLM writing tools became widely available in late 2022, the researchers found that ‘the appearance of LLMs led to an abrupt increase in the frequency of certain style words’ that was ‘unprecedented in both quality and quantity.’

“To measure these vocabulary changes, the researchers analyzed 14 million paper abstracts published on PubMed between 2010 and 2024, tracking the relative frequency of each word as it appeared across each year. They then compared the expected frequency of those words (based on the pre-2023 trend line) to the actual frequency of those words in abstracts from 2023 and 2024, when LLMs were in widespread use.

“The results found a number of words that were extremely uncommon in these scientific abstracts before 2023 that suddenly surged in popularity after LLMs were introduced. The word ‘delves,’ for instance, shows up in 25 times as many 2024 papers as the pre-LLM trend would expect; words like ‘showcasing’ and ‘underscores’ increased in usage by nine times as well. Other previously common words became notably more common in post-LLM abstracts: The frequency of ‘potential’ increased by 4.1 percentage points, ‘findings’ by 2.7 percentage points, and ‘crucial’ by 2.6 percentage points, for instance.

These kinds of changes in word use could happen independently of LLM usage, of course — the natural evolution of language means words sometimes go in and out of style.

“However, the researchers found that, in the pre-LLM era, such massive and sudden year-over-year increases were only seen for words related to major world health events. …

“In the post-LLM period, though, the researchers found hundreds of words with sudden, pronounced increases in scientific usage that had no common link to world events. … The words with a post-LLM frequency bump were overwhelmingly ‘style words’ like verbs, adjectives, and adverbs (a small sampling: ‘across, additionally, comprehensive, crucial, enhancing, exhibited, insights, notably, particularly, within’). …

“The pre-2023 set of abstracts acts as its own effective control group to show how vocabulary choice has changed overall in the post-LLM era.

“By highlighting hundreds of so-called ‘marker words’ that became significantly more common in the post-LLM era, the telltale signs of LLM use can sometimes be easy to pick out. Take this example abstract line called out by the researchers, with the marker words highlighted: ‘A comprehensive grasp of the intricate interplay between […] and […] is pivotal for effective therapeutic strategies.’

“After doing some statistical measures of marker word appearance across individual papers, the researchers estimate that at least 10 percent of the post-2022 papers in the PubMed corpus were written with at least some LLM assistance. The number could be even higher, the researchers say, because their set could be missing LLM-assisted abstracts that don’t include any of the marker words they identified. …

“Papers authored in countries like China, South Korea, and Taiwan showed LLM marker words 15 percent of the time, suggesting ‘LLMs might … help non-natives with editing English texts, which could justify their extensive use.’ On the other hand, the researchers offer that native English speakers ‘may [just] be better at noticing and actively removing unnatural style words from LLM outputs,’ thus hiding their LLM usage from this kind of analysis.”

More at Ars Technica via Wired, here.

Read Full Post »

Machover Weighs In On AI

Posted in Uncategorized, tagged , , , , , on November 15, 2023| Leave a Comment »

Photo: Sam Odgden via Chamber Music America.
Composer Tod Machover.

With all the furor about artificial intelligence, Rebecca Schmid decided to check in with MIT’s Tod Machover, “a pioneer of the connections between classical music and computers.” Their conversation about how AI applies to music appears on the Chamber Music America website.

“Sitting at his home in Waltham, Massachusetts, the composer Tod Machover speaks with the energy of someone half his 69 years as he reflects on the evolution of digital technology toward the current boom in artificial intelligence.

“ ‘I think the other time when things moved really quickly was 1984,’ he says — the year when the personal computer came out. Yet he sees this moment as distinct. ‘What’s going on in A.I. is like a major, major difference, conceptually, in how we think about music and who can make it.’

“Perhaps no other figure is better poised than Machover to analyze A.I.’s practical and ethical challenges. The son of a pianist and computer graphics pioneer, he has been probing the interface of classical music and computer programming since the 1970s.

“As the first Director of Musical Research at the then freshly opened Institut de Recherche et Coordination Acoustique/Musique (I.R.C.A.M.) in Paris, he was charged with exploring the possibilities of what became the first digital synthesizer while working closely alongside Pierre Boulez.

“In 1987, Machover introduced Hyperinstruments for the first time in his chamber opera VALIS, a commission from the Pompidou Center in Paris. This technology incorporates innovative sensors and A.I. software to analyze the expression of performers, allowing changes in articulation and phrasing to turn, in the case of VALIS, keyboard and percussion soloists into multiple layers of carefully controlled sound.

“Machover had helped to launch the M.I.T. Media Lab two years earlier in 1985, and now serves as both Muriel R. Cooper Professor of Music and Media and director of the Lab’s Opera of the Future group. …

“Machover emphasizes the need to blend the capabilities of [AI] technology with the human hand. For his new stage work, Overstory Overture, which premiered last March at Lincoln Center, he used A.I. as a multiplier of handmade recordings to recreate the sounds of forest trees ‘in underground communication with one another.’

“Machover’s ongoing series of ‘City Symphonies,’ for which he involves the citizens of a given location as he creates a sonic portrait of their hometown, also uses A.I. to organize sound samples. Another recent piece, Resolve Remote, for violin and electronics, deployed specially designed algorithms to create variations on acoustic violin. …

“Machover has long pursued his interest in using technology to involve amateurs in musical processes. His 2002 Toy Symphony allows children to shape a composition, among other things, by means of ‘beat bugs’ that generate rhythms. This work, in turn, spawned the Fisher-Price toy Symphony Painter and has been customized to help the disabled imagine their own compositions. …

“Rebecca Schmid: How is the use of A.I. a natural development from what you began back in the 1970s, and what is different?
“Tod Machover: There are lots of things that could only be done with physical instruments 30 years ago that are now done in software: you can create amazing things on a laptop. But what’s going on in A.I. is like a major, major difference, conceptually, in how we think about music and who can make it.

“One of my mentors and heroes is Marvin Minsky, who was one of the founders of A.I., and a kind of music prodigy. And his dream for A.I. was to really figure out how the mind works. He wrote a famous book called The Society of Mind in the mid-eighties based on an incredibly radical, really beautiful theory: that your mind is a group of committees that get together to solve simple problems, with a very precise description of how that works. He wanted a full explanation of how we feel, how we think, how we create — and to build computers modeled on that.

“Little by little, A.I. moved away from that dream, and instead of actually modeling what people do, started looking for techniques that create what people do without following the processes at all. A lot of systems in the 1980 and 1990s were based on pretty simple rules for a particular kind of problem, like medical diagnosis. You could do a pretty good job of finding out some similarities in pathology in order to diagnose something. But that system could never figure out how to walk across the street without getting hit by a car. It had no general knowledge of the world.

“We spent a lot of time in the seventies, eighties, and nineties trying to figure out how we listen — what goes on in the brain when you hear music, how you can have a machine listen to an instrument — to know how to respond. A lot of the systems which are coming out now don’t do that at all. They don’t pretend to be brains. Some of the most kind of powerful systems right now, especially ones generating really crazy and interesting stuff, look at pictures of the sound — a spectrogram, a kind of image processing. I think it’s going to reach a limit because it doesn’t have any real knowledge of what’s there. So, there’s a question of, what does it mean and how is it making these decisions?

“What systems have you used successfully in your work?
“One is R.A.V.E., which comes from I.R.C.A.M. and was originally developed to analyze audio, especially live audio, so that you can reconstruct and manipulate it. The voice is a really good example. Ever since the 1950s, people have been doing live processing of singing. The problem is that it’s really hard to analyze everything that’s in the voice: The pitch and spectrum are changing all the time.

“What you really want to do is be able to understand what’s in the voice, pull it apart and then have all the separate elements so that you can tune and tweak things differently on the other side. And that’s what R.A.V.E. was invented to do. It’s an A.I. analysis of an acoustic signal. It reconstructs it in some form, and then ideally it comes out the other side sounding exactly like it did originally, but now it’s got all these handles so that I can change the pitch without changing the timbre. And it works pretty well for that. You can have it as an accompanist, or your own voice can accompany you. It can change pitch and sing along. And it can sing things that you never sang because it understands your voice. …

“The great thing about A.I. models now is that you can use them not just to make a variation in the sound, but also a variation in what’s being played. So, if you think about early electronic music serving to kind of color a sound — or add a kind of texture around the sound, but being fairly static — with this, if you tweak it properly, it’s a kind of complex variation closely connected to what comes in but not exactly the same. And it changes all the time, because every second the A.I. is trying to figure out, How am I going to match this? How far am I going to go? Where in the space am I? You can think of it as a really rich way of transforming something or creating a kind of dialogue with the performer.” Lots more at Chamber Music America, here. No firewall.

I myself have posted about the composer a few times: for example, here (“Tod Machover,” 2012); here (“Stanford’s Laptop Orchestra,” 2018); and here (“Symphony of the Street,” 2017).

“AI Finished My Story. Does It Matter?” at Wired, here, offers additional insight.

Read Full Post »

AI Tackles Ancient Scripts

Posted in Uncategorized, tagged , , , , on December 18, 2020| 6 Comments »

Photo: Ismoon
The earliest recognized form of “written” communication may have been small bits of clay, called tokens. This charming one is from the Indus Valley.

Nowadays, one reads almost too much about artificial intelligence, AI. I myself have an ever increasing list of things I’d rather not have AI managing for me. But using it to translate ancient texts is one application that seems to make perfect sense.

Ruth Schuster writes at Haaretz, “Understanding texts written using an unknown system in a tongue that’s been dead for thousands of years is quite the challenge. Reconstructing missing bits of the ancient text is even harder. …

“Filling in missing text starts with being able to read and understand the original text. That requires much donkey work. Now an Israeli team led by Shai Gordin at Ariel University in the West Bank has reinvented the donkey in digital form, harnessing artificial intelligence to help complete fragmented Akkadian cuneiform tablets.

“Their paper, ‘Restoration of Fragmentary Babylonian Texts Using Recurrent Neural Networks,’ was published in the Proceedings of the National Academy of Sciences in September.

“ ‘Neural networks’ … means software inspired by biological nervous systems. The concept dates back more than 70 years. … The base concept is to teach machines to learn, think and make decisions. In this case, the computer decides on the plausible completion of missing text. …

“Gordin and the team feed their machine transliterations of the extant Babylonian texts i.e., what the text would have sounded like.

“Then what? When it comes to missing bits in a papyrus or tablet, humans can intuit that “’…ow is your moth…’ isn’t a query into the well-being of your mothball.

With machines, it’s all about mathematics and probabilities based on knowledge gained so far. …

“It may have been trading that inspired the earliest recognized form of communication: ‘pseudo-writing’ on small bits of clay in Mesopotamia around 7,000 years ago. The clay bits, called tokens, were shaped into simplistic imagery such as a cow or other ancient commodities. …

“Then we start seeing abstract signs; repetitive strokes or depressions are interpreted as numbers (price, perhaps); and possibly also personal names, using the first sounds of different imprints to put together words you can’t draw. …

“Anyway, after pseudo-writing came proto-writing: figurative proto-cuneiform inscribed on tablets, which arose about 5,500 years ago in the city of Uruk. … Within mere centuries, proto-cuneiform evolved to become increasingly schematic and Sumer was apparently where it happened, [Gordin] says. And figurative hieroglyphic script began to appear in ancient Egypt at about the same time, about 5,000 years ago. …

“By the time cuneiform became a thing, writing had passed the stage of ‘Sheep : four : Yerachmiel’ and reached the stage of official records, letters and formulaic recounts of the wondrousness of the ruler. …

“For cuneiform, we have the gargantuan multilingual text at Behistun, Iran. Darius the Great had his exploits described in three different cuneiform scripts. [The] Behistun text was monumental: 15 meters (49 feet) high by 25 meters wide, and 100 meters up a cliff on the road connecting Babylon and Ecbatana, all to describe how Darius vanquished Gaumata and other foes. …

“And over decades, linguists slowly interpreted the languages of Babylon and Assyria, thanks to Darius’ monumental ego. …

“Interpreting a dead language is a mathematical game, Gordin says. … Neural networks are a computerized model that can understand text. How? They turn each symbol or word into a number, he explains. …

“When humans reconstruct missing text, their interpretation may be subjective. To be human is to err with bias, and quantifying the likely accuracy of the completion is impossible. Enter the machine. …

“The machine proved capable of identifying sentence structures – and did better than expected in making semantic identifications on the basis of context-based statistical inference, Gordin says. Its talents were further deduced by designing a completion test, in which the machine-learning model had to answer a multiple-choice question: which word fits in the blank space of a given sentence.”

Not sure how many readers are into that kind of thing, but I do find it intriguing. More at Haaretz, here.

Read Full Post »