In brief AI text-to-image generation models are all the rage right now. You give them a simple description of a scene, such as "a vulture typing on a laptop," and they come up with an illustration that resembles that description.

That's the theory, anyway. But developers who have special access to OpenAI's text-to-image engine DALL·E 2 have found all sorts of weird behaviors – including what may be a hidden, made-up language.

Giannis Daras, a PhD student at the University of Texas at Austin shared artwork produced by DALL·E 2 given the input: "Apoploe vesrreaitais eating Contarra ccetnxniams luryca tanniounons" – a phrase that makes no sense to humans. But to the machine, it seemed to generate images of birds eating bugs consistently. Daras made the following claim:

DALL·E 2 has a secret language. "Apoploe vesrreaitais" means birds. "Contarra ccetnxniams luryca tanniounons" means bugs or pests. The prompt: "Apoploe vesrreaitais eating Contarra ccetnxniams luryca tanniounons" gives images of birds eating bugs. A thread (1/n)🧵 pic.twitter.com/VzWfsCFnZo

The PhD student said the examples showed DALL·E 2 has an understanding of some mysterious language, in which "Apoploe vesrreaitais" means birds and "Contarra ccetnxniams luryca tanniounons" means bugs or pests, apparently. But another researcher named Benjamin Hilton tested Daras's claims, adding the words "3D render" to the same input prompt. Instead of just birds eating bugs, Hilton got pictures of "sea-related things." The prompt "Contarra ccetnxniams luryca tanniounons" on its own also generated images of random animals and not bugs:

Let's start with some of the basic claims. 1) @giannis_daras says "Contarra ccetnxniams luryca tanniounons" means bugs or pests. This just seems wrong. Here's what I get if I put "Contarra ccetnxniams luryca tanniounons" into DALL-E – lots of different animals. (2/15) pic.twitter.com/RGHeRw1pmb

Tweaking the inputs by adding random words to the original "Apoploe vesrreaitais eating Contarra ccetnxniams luryca tanniounons" prompt makes DALL·E 2 produce strange images of grandmas, beetles, or vegetables. Hilton believes that it shows the model doesn't have a secret understanding of some unknown language, but instead demonstrates the random nature of AI. Why exactly DALL·E 2 associates those images with the gibberish inputs remains unclear.

Reports of a racially biased algorithm used to help social workers decide whether to investigate families for child neglect has prompted officials in Oregon to drop a similar tool they have used since 2018.

An investigation led by the Associated Press found that the automated screening tool used in Allegheny County, Pennsylvania, was more likely to recommend black children for  "mandatory" neglect investigations than white children. Child welfare workers in Oregon used similar software to calculate risk scores for different families, higher scores meaning parents or guardians were more likely mean to be scrutinized for neglecting their children. 

But now officials have decided to scrap the tool, and will stop using it by the end of this month. Jake Sunderland, a spokesman for Oregon's Department of Human Services, told AP the algorithm would "no longer be necessary." He did not explain why it had been dropped. 

Oregon won't be getting rid of automated screening tools completely, however. Its existing algorithm will be replaced by another one called the Structured Decision-Making model. Another tool that helps decide whether foster care children should be reunited with their families has been paused. 

Meta's VP of AI is leaving his role leading Zuck & Co.'s machine learning research team, FAIR, as the org reshuffles internal teams. 

After working at the social media biz for over four years, Jerome Pesenti will exit in mid-June. "We are grateful for the incredible work Jerome has done over the past 4+ years in building, leading, and scaling a world-class AI function for Meta," Andrew Bosworth, the company's CTO, said in an announcement this week. 

"FAIR will continue to have incredibly strong leadership in place with Joelle Pineau, Antoine Bordes, and Yann LeCun."

As part of the reshuffle, Meta is breaking up teams at FAIR so they can join other departments on the product side of the business. The Responsible AI unit, for example, will be folded under the Social Impact team, and the AI4AR team will join with the XR team in Reality Labs.

"More centralized approaches run into their limits when the last mile proves to be too far for downstream teams to close the gap. With this new team structure, we are excited to push the boundaries of what AI can do and use it to create new features and products for billions of people," Bosworth concluded.

What happens when you're defamed by an internet giant's code? Can you sue? Whom do you sue? Does it matter if America's First Amendment is in play?

Cybersecurity guru Marcus Hutchins, best known for halting the WannaCry ransomware epidemic, is sometimes wrongly named by Google's search engine as having created the so-called crypto-worm. Hutchins said googling his name used to return an automatically generated passage of text in the search results that incorrectly stated he was the one who developed the WannaCry malware.

Being automatically mislabeled by Google in search results has a noticeable effect, he said.

"Ever since Google started featuring these SERPs [search engine results pages], I've gotten a huge spike in hate comments and even threats based on me creating WannaCry," he told TechCrunch.

"The timing of my legal case gives the impression that the FBI suspected me but a quick [Google search] would confirm that's not the case. Now there's all kinds of SERP results which imply I did, confirming the searcher's suspicious and it's caused rather a lot of damage to me."

Hutchins was arrested, charged and spared jail time for developing another strain of malware, unrelated to WannaCry, when he was a teenager.

Google's language model doesn't quite know the difference. A spokesperson said the search engine giant had removed the generated text for Hutchins so that specific problem shouldn't crop up again.

But what if false defamatory remarks affect someone less famous than Hutchins? What if they have a real negative impact on someone's future opportunities? There is currently little regulation around AI-generated content, and it's not clear how well this fits in around today's libel laws. ®

AI is killing the planet. Wait, no – it's going to save it. According to Hewlett Packard Enterprise VP of AI and HPC Evan Sparks and professor of machine learning Ameet Talwalkar from Carnegie Mellon University, it's not entirely clear just what AI might do for – or to – our home planet.

Speaking at the SixFive Summit this week, the duo discussed one of the more controversial challenges facing AI/ML: the technology's impact on the climate.

"What we've seen over the last few years is that really computationally demanding machine learning technology has become increasingly prominent in the industry," Sparks said. "This has resulted in increasing concerns about the associated rise in energy usage and correlated – not always cleanly – concerns about carbon emissions and carbon footprint of these workloads."

HPE is lifting the lid on a new AI supercomputer – the second this week – aimed at building and training larger machine learning models to underpin research.

Based at HPE's Center of Excellence in Grenoble, France, the new supercomputer is to be named Champollion after the French scholar who made advances in deciphering Egyptian hieroglyphs in the 19th century. It was built in partnership with Nvidia using AMD-based Apollo computer nodes fitted with Nvidia's A100 GPUs.

Champollion brings together HPC and purpose-built AI technologies to train machine learning models at scale and unlock results faster, HPE said. HPE already provides HPC and AI resources from its Grenoble facilities for customers, and the broader research community to access, and said it plans to provide access to Champollion for scientists and engineers globally to accelerate testing of their AI models and research.

A prankster researcher has trained an AI chatbot on over 134 million posts to notoriously freewheeling internet forum 4chan, then set it live on the site before it was swiftly banned.

Yannic Kilcher, an AI researcher who posts some of his work to YouTube, called his creation "GPT-4chan" and described it as "the worst AI ever". He trained GPT-J 6B, an open source language model, on a dataset containing 3.5 years' worth of posts scraped from 4chan's imageboard. Kilcher then developed a chatbot that processed 4chan posts as inputs and generated text outputs, automatically commenting in numerous threads.

Netizens quickly noticed a 4chan account was posting suspiciously frequently, and began speculating whether it was a bot.

Updated Australia's federal police and Monash University are asking netizens to send in snaps of their younger selves to train a machine-learning algorithm to spot child abuse in photographs.

Researchers are looking to collect images of people aged 17 and under in safe scenarios; they don't want any nudity, even if it's a relatively innocuous picture like a child taking a bath. The crowdsourcing campaign, dubbed My Pictures Matter, is open to those aged 18 and above, who can consent to having their photographs be used for research purposes.

All the images will be amassed into a dataset managed by Monash academics in an attempt to train an AI model to tell the difference between a minor in a normal environment and an exploitative, unsafe situation. The software could, in theory, help law enforcement better automatically and rapidly pinpoint child sex abuse material (aka CSAM) in among thousands upon thousands of photographs under investigation, avoiding having human analysts inspect every single snap.

As compelling as the leading large-scale language models may be, the fact remains that only the largest companies have the resources to actually deploy and train them at meaningful scale.

For enterprises eager to leverage AI to a competitive advantage, a cheaper, pared-down alternative may be a better fit, especially if it can be tuned to particular industries or domains.

That’s where an emerging set of AI startups hoping to carve out a niche: by building sparse, tailored models that, maybe not as powerful as GPT-3, are good enough for enterprise use cases and run on hardware that ditches expensive high-bandwidth memory (HBM) for commodity DDR.

The US Copyright Office and its director Shira Perlmutter have been sued for rejecting one man's request to register an AI model as the author of an image generated by the software.

You guessed correct: Stephen Thaler is back. He said the digital artwork, depicting railway tracks and a tunnel in a wall surrounded by multi-colored, pixelated foliage, was produced by machine-learning software he developed. The author of the image, titled A Recent Entrance to Paradise, should be registered to his system, Creativity Machine, and he should be recognized as the owner of the copyrighted work, he argued.

(Owner and author are two separate things, at least in US law: someone who creates material is the author, and they can let someone else own it.)

IBM's self-sailing Mayflower Autonomous Ship (MAS) has finally crossed the Atlantic albeit more than a year and a half later than planned. Still, congratulations to the team.

That said, MAS missed its target. Instead of arriving in Massachusetts – the US state home to Plymouth Rock where the 17th-century Mayflower landed – the latest in a long list of technical difficulties forced MAS to limp to Halifax in Nova Scotia, Canada. The 2,700-mile (4,400km) journey from Plymouth, UK, came to an end on Sunday.

The 50ft (15m) trimaran is powered by solar energy, with diesel backup, and said to be able to reach a speed of 10 knots (18.5km/h or 11.5mph) using electric motors. This computer-controlled ship is steered by software that takes data in real time from six cameras and 50 sensors. This application was trained using IBM's PowerAI Vision technology and Power servers, we're told.

In brief Miscreants can easily steal someone else's identity by tricking live facial recognition software using deepfakes, according to a new report.

Sensity AI, a startup focused on tackling identity fraud, carried out a series of pretend attacks. Engineers scanned the image of someone from an ID card, and mapped their likeness onto another person's face. Sensity then tested whether they could breach live facial recognition systems by tricking them into believing the pretend attacker is a real user.

So-called "liveness tests" try to authenticate identities in real-time, relying on images or video streams from cameras like face recognition used to unlock mobile phones, for example. Nine out of ten vendors failed Sensity's live deepfake attacks.

DALL·E 2 may have to cede its throne as the most impressive image-generating AI to Google, which has revealed its own text-to-image model called Imagen.

Like OpenAI's DALL·E 2, Google's system outputs images of stuff based on written prompts from users. Ask it for a vulture flying off with a laptop in its claws and you'll perhaps get just that, all generated on the fly.

A quick glance at Imagen's website shows off some of the pictures it's created (and Google has carefully curated), such as a blue jay perched on a pile of macarons, a robot couple enjoying wine in front of the Eiffel Tower, or Imagen's own name sprouting from a book. According to the team, "human raters exceedingly prefer Imagen over all other models in both image-text alignment and image fidelity," but they would say that, wouldn't they.

The future of high-performance computing will be virtualized, VMware's Uday Kurkure has told The Register.

Kurkure, the lead engineer for VMware's performance engineering team, has spent the past five years working on ways to virtualize machine-learning workloads running on accelerators. Earlier this month his team reported "near or better than bare-metal performance" for Bidirectional Encoder Representations from Transformers (BERT) and Mask R-CNN — two popular machine-learning workloads — running on virtualized GPUs (vGPU) connected using Nvidia's NVLink interconnect.

NVLink enables compute and memory resources to be shared across up to four GPUs over a high-bandwidth mesh fabric operating at 6.25GB/s per lane compared to PCIe 4.0's 2.5GB/s. The interconnect enabled Kurkure's team to pool 160GB of GPU memory from the Dell PowerEdge system's four 40GB Nvidia A100 SXM GPUs.

The Register - Independent news and views for the tech community. Part of Situation Publishing

Biting the hand that feeds IT © 1998–2022