Tag Archives: Artificial intelligence

BuzzFeed says it will use AI to help create content, stock jumps 150%


New York
CNN
 — 

BuzzFeed said Thursday that it will work with ChatGPT creator OpenAI to use artificial intelligence to help create content for its audience, marking a milestone in how media companies implement the new technology into their businesses.

Jonah Peretti, the company’s co-founder and chief executive, told employees in a memo that they can expect “AI inspired content” to “move from an R&D stage to part of our core business.”

Peretti elaborated that the technology will be used to create quizzes, help with brainstorming, and assist in personalizing content to its audience. BuzzFeed, for now, will not use artificial intelligence to help write news stories, a spokesperson told CNN.

“To be clear, we see the breakthroughs in AI opening up a new era of creativity that will allow humans to harness creativity in new ways with endless opportunities and applications for good,” Peretti said. “In publishing, AI can benefit both content creators and audiences, inspiring new ideas and inviting audience members to co-create personalized content.”

“When you see this work in action it is pretty amazing,” Peretti added, vowing to “lead the future of AI-powered content.”

The news sent BuzzFeed’s sagging stock skyrocketing more than 150% in trading Thursday to more than $2 a share.

Media industry leaders have increasingly said that artificial intelligence will revolutionize their businesses.

While BuzzFeed is the biggest digital content creator to move to implement OpenAI’s technology into its business, some other outlets have taken similar steps.

CNET recently used an artificial intelligence tool to help write stories. But the process did not go smoothly, with a number of articles ultimately requiring corrections.

In a note published online Wednesday, CNET Editor-In-Chief Connie Guglielmo apologized for the errors and said new processes had been put in place to prevent them in the future.

But, Guglielmo said, the outlet will not shy away from using artificial intelligence moving forward.

“The process may not always be easy or pretty, but we’re going to continue embracing it – and any new tech that we believe makes life better,” Guglielmo wrote.

The Associated Press also began using artificial intelligence to automate news stories nearly a decade ago.

Read original article here

After ChatGPT and DALL-E, meet VALL-E – the text-to-speech AI that can mimic anyone’s voice

Last year saw the emergence of artificial intelligence tools (AI) that can create images, artwork, or even video with a text prompt.

There were also major steps forward in AI writing, with OpenAI’s ChatGPT causing widespread excitement – and fear – about the future of writing.

Now, just a few days into 2023, another powerful use case for AI has stepped into the limelight – a text-to-voice tool that can impeccably mimic a person’s voice.

Developed by Microsoft, VALL-E can take a three-second recording of someone’s voice, and replicate that voice, turning written words into speech, with realistic intonation and emotion depending on the context of the text.

Trained with 60,000 hours worth of English speech recordings, it can deliver a speech in a “zero-shot situation,” which means without any prior examples or training in a specific context or situation.

Introducing VALL-E in a paper published by Cornell University, the developers explained that the recording data consisted of more than 7,000 unique speakers.

The team say their Text To Speech system (TTS) used hundreds of times more data than the existing TTS systems, helping them to overcome the zero-shot issue.

The tool is not currently available for public use – but it does throw up questions about safety, given it could feasibly be used to generate any text coming from anybody’s voice.

Microsoft betting big on AI

Its creators have, however, provided a demo, showcasing a number of three-second speaker prompts and a demonstration of the text-to-speech in action, with the voice correctly mimicked.

Alongside the speaker prompt and VALL-E’s output, you can compare the results with the “ground truth” – the actual speaker reading the prompt text – and the “baseline” result from current TTS technology.

Microsoft has invested heavily in AI and is one of the backers of OpenAI, the company behind ChatGPT and DALL-E, a text-to-image or art tool.

The software giant invested $1 billion (€930 million) in OpenAI in 2019, and a report this week on semafor.com stated it was looking at investing another $10 billion (€9.3 billion) in the company.

Read original article here

10 moments in 2022 straight out of a sci-fi movie

Sign up for CNN’s Wonder Theory science newsletter. Explore the universe with news on fascinating discoveries, scientific advancements and more.



CNN
 — 

From a spacecraft the size of a refrigerator plowing into an asteroid (deliberately) to a helicopter trying to catch a rocket plummeting back to Earth, 2022 offered surreal moments in space that could have been ripped from the pages of a science fiction movie script.

Among the memorable events were billionaires mapping out plans to explore the cosmos and scientists attempting to find answers to perplexing questions, only to discover deeper mysteries.

Researchers managed to grow plants in lunar soil for the first time, while engineers successfully tested an inflatable heat shield that could land humans on Mars. And scientists determined that a rare interstellar meteor crashed into Earth nearly a decade ago.

Here’s a look back at 10 times space travel and exploration felt more like a plot from a Hollywood movie than real life.

A NASA spacecraft intentionally slammed into Dimorphos, a small asteroid that orbits a larger space rock named Didymos. While this collision seemed like something out of the 1998 movie “Armageddon,” the Double Asteroid Redirection Test was a demonstration of deflection technology — and the first conducted on behalf of planetary defense.

Many tuned in on September 26 to watch as the surface of Dimorphos came into view for the first time, with DART’s cameras beaming back live imagery. The view ended after the spacecraft collided with the asteroid, but images captured by space telescopes and an Italian satellite provided dramatic photos of the aftermath.

The DART mission marked the first time humanity intentionally changed the motion of a celestial object in space. The spacecraft altered the moonlet asteroid’s orbit by 32 minutes. Neither Dimorphos nor Didymos pose a threat to Earth, but the double-asteroid system was a perfect target to test deflection technology.

Fast radio bursts in space have intrigued astronomers since their 2007 discovery, but a mysterious radio burst with a pattern similar to a heartbeat upped the ante this year.

Fast radio bursts, or FRBs, are intense, millisecond-long bursts of radio waves with unknown origins — which only fuels speculation that their cause is more alien than cosmic.

Astronomers estimate that the “heartbeat signal” came from a galaxy roughly 1 billion light-years away, but the location and cause of the burst are unknown.

Additionally, astronomers also detected a powerful radio wave laser, known as a megamaser, and a spinning celestial object releasing giant bursts of energy unlike anything they had ever seen before.

Speaking of strange objects, astronomers made a new leap forward in understanding odd radio circles, or ORCs. No, they aren’t the goblinlike humanoids from “The Lord of the Rings” books, but these fascinating objects have baffled scientists since their discovery in 2020.

The space rings are so massive that they each measure about 1 million light-years across — 16 times bigger than our Milky Way galaxy. Astronomers believe it takes the circles 1 billion years to reach their maximum size, and they are so large they have expanded past other galaxies.

Astronomers took a new detailed photo of odd radio circles using the South African Radio Astronomy Observatory’s MeerKAT telescope, narrowing down the possible theories that might explain these celestial oddballs.

Black holes are known for behaving badly and shredding stars — so astronomers using the Hubble Space Telescope were surprised when they saw a black hole fueling star birth.

Their observation revealed a gaseous umbilical cord stretching from a black hole at the center of a dwarf galaxy to a stellar nursery where stars are born. The stream of gas provided by the black hole triggered a fireworks show of star birth as it interacted with the cloud, which led to a cluster of forming stars.

This year, astronomers also captured an image of the supermassive black hole lurking at the center of our galaxy, and Hubble spied a lone black hole wandering the Milky Way. And X-ray signals from black holes were converted into eerie sounds we won’t soon forget.

Rocket Lab, a US-based company that launches out of New Zealand, is trying to figure out a way to recapture its rocket boosters as they tumble down toward Earth after launch. In 2022, the company made two attempts to deploy a helicopter with a hook attachment. The wild spectacle is all part of Rocket Lab’s plans to save money by recovering and reusing rocket parts after they vault satellites to space.

The first attempt in May appeared to go as planned when the helicopter snagged a booster. But the pilots made the decision to drop the rocket part due to safety concerns.

On the second attempt, the rocket never came into view, and pilots confirmed the booster wouldn’t be returning to the factory dry. In a tweet, the company reported there was a data loss issue during the rocket’s reentry.

NASA flew its first virtual assistant on a moon mission with the space agency’s historic Artemis I flight — a version of Amazon’s Alexa.

While not exactly reminiscent of HAL 9000, the antagonistic voice assistant in “2001: A Space Odyssey,” the decision did spark plenty of facetious comparisons.

The Artemis I mission was uncrewed, but NASA’s ground control teams used the voice assistant, called Callisto, to control cabin lighting and play music during the journey. It did not have the ability to open or close doors, for the record.

Artemis I was just a test mission, and NASA is still evaluating how the voice recognition system may be included on future missions.

Japanese fashion mogul Yusaku Maezawa picked eight passengers who he said will join him on a trip around the moon, powered by SpaceX’s yet-to-be-flown Starship spacecraft. The group includes American DJ Steve Aoki and popular space YouTuber Tim Dodd, better known as the Everyday Astronaut.

The mission, called Dear Moon, was first announced in 2018 with the intention of flying by 2023. Maezawa initially aimed to take a group of artists with him on a six-day trip around the moon but later announced he had expanded his definition of an “artist.” Instead, Maezawa announced in a video last year that he would be open to people from all walks of life as long as they viewed themselves as artists.

Separately, millionaire Dennis Tito — who became the first person to pay his way to the International Space Station in the early 2000s — made his own lunar travel plans with SpaceX.

Chunks of space debris were reportedly found on farmland in Australia’s Snowy Mountains, and NASA and authorities confirmed that the objects were likely scraps of hardware from a SpaceX Dragon capsule intentionally jettisoned as the spacecraft reentered Earth’s atmosphere in May 2021.

It’s common for space debris to fall to Earth. But it’s far less common for the objects to wind up on land since most space garbage is discarded in the ocean.

Perhaps among the most unique space start-ups in the world, SpinLaunch aims to whip satellites around in a vacuum-sealed chamber and toss them into space rather than put them on a rocket.

The company began testing a scaled-down version of its technology last year, but things ramped up in 2022. SpinLaunch notched its 10th test flight in October.

There’s a science fiction connection as well. SpinLaunch founder Jonathan Yaney cites the work of Jules Verne — the “Journey to the Center of Earth” writer who died more than 50 years before the first satellite traveled to space — as the inspiration for SpinLaunch.

It’s not clear whether the company’s technology will ever come to fruition. But in the meantime, this group will be in the New Mexico desert attempting to bring art to life.

If it wasn’t surreal enough watching Amazon billionaire Jeff Bezos and other celebrities travel to space on his self-funded, suborbital rocket last year, hearing that the rocket exploded a little more than a year later over West Texas — albeit on a trip without any passengers — was a harrowing moment that brought home the adage “space is hard.” However, the crew capsule, which was carrying science projects and other inanimate payloads on September 12, was able to land successfully.

“The capsule landed safely and the booster impacted within the designated hazard area,” the Federal Aviation Administration said in a September statement. Bezos’ Blue Origin has been in limbo since and has not returned to flight.

And with Richard Branson’s Virgin Galactic still grounded, neither of the companies spearheading suborbital space tourism last year are conducting routine flights.



Read original article here

Artists Protest As ArtStation Allows AI-Generated Art On Site

ArtStation is probably the most important website on the whole internet for professional artists, especially those working in entertainment fields like video games (most of our Fine Art links, for example, point there). Which is why the site’s continued allowance of AI-generated imagery has become a point of contention with its users.

The technology, which is rotten to its core, is of particular concern to a community who make a living creating art, and as such should also be a concern to the companies responsible for owning and hosting that community. But as of today, ArtStation has no policy directly restricting the hosting or display of AI-generated imagery on the site, which has led to repeated instances where images made by computers, and not humans, have floated to the top of ArtStation’s “Explore” section, its most popular means of showcasing the work of artists.

That is, understandably, pissing a lot of people off. Indeed, over the last 24 hours so many artists have become so incensed by the site’s allowance of AI-generated imagery that they’ve begun spamming their portfolios, with a protest sparked by illustrator Nicholas Kole and costume designer Imogen Chayes resulting in ArtStation’s front page looking like this at time of writing:

It’s just the same image, originally created by Alexander Nanitchkov and saying “No To AI Generated Images”, pasted over and over again by hundreds of artists:

These artists are right to be upset! The rapidly-encroaching practice of AI-generated imagery is going to trash all kinds of websites, but to allow it on a site specifically designed to showcase the work of talented human artists is an especially bad look.

“ArtStation’s content guidelines do not prohibit the use of AI tools in the process of creating artwork that is shared with the community”, a spokesperson for Epic Games, the owners of ArtStation, tells Kotaku. “That said, ArtStation is a portfolio platform designed to elevate and celebrate originality powered by a community of artists. Users’ portfolios should only feature artwork that they create, and we encourage users to be transparent in the process. Our content guidelines are here.”

The profile of a user submitting AI-generated imagery to ArtStation earlier today
Screenshot: ArtStation

While that’s an expected response given the prevalence of AI-generated imagery currently on the site, and the apparent lack of moderation involved in letting them stay up, Epic also say they “do not make any agreements with companies allowing them to scrape content on our website. If AI companies are doing this without permission and beyond purely academic use (where copyright fair use may apply), they may be infringing the rights of ArtStation creators.”

Epic also say they are “in the process of giving ArtStation users more control over how their work is shared and labeled, and we will provide more details in the near future.”

While that veiled legal threat is perhaps a sign that Epic aren’t quite as cool with the practice as it seems, and word that user controls are coming in the “near future” is promising to an extent, that doesn’t change the fact that ArtStation user’s portfolios have already been fed to these AIs, and that it won’t do anything in the short term to stop AI-generated images from encroaching on a website that’s supposed to be showcasing the best in human art.

For now, the best way to detect AI-generated imagery and ignore it (or even better, to report it) is the same way it has been for the past few months: always ask to see the fingers.



Read original article here

Adobe Embraces AI-Generated Stock Art, Can Get In The Bin

Adobe used to be known as the company that made Acrobat and PhotoShop. Addobe is increasingly becoming known, however, as one of the great digital grifters of the modern age.

From its shonky subscription models to making people pay for certain colours in PhotoShop (big shout out to Pantone there as well), the company is, like so many others in these tumultuous times, more concerned with growing its bottom line at any cost than it is in taking a moment to consider the needs of its users, or the consequences of its actions.

I’m bringing this up today because, less than a week after forcing people to check they weren’t reading an Onion story when learning about the colours thing, the company has announced that it is embracing AI art, which is not only an enormous grift, but also a serious threat to the livelihoods of artists around the world, big and small.

I’ve made my feelings about AI very clear on this website already—I wrote this feature back in August interviewing a range of video game and entertainment industry artists—and think it sucks not just because it’s a threat to artists, but to art. While people’s jobs are of course important, we’re not just talking about cotton gins here, and how this is in many ways a labour v capital breakdown; we’re talking about a process that is encroaching on a fundamentally human pastime and creative pursuit.

Machines don’t make art. They’re machines! They’re just making an approximated casserole out of human art that has been fed into it, in the vast amount of cases without credit or compensation. As Dan Sheehan says in his fantastic piece Art In The Age Of Optimization, AI art isn’t about art, it’s merely “a technology that clearly exists to remove the human element from the process of artistic expression”.

Anyway! Last week Adobe dropped an announcement saying that AI-generated art was going to be made available as part of the company’s vast library of stock images, going so far as to say the field is “amplifying human creativity”. The company boldly says, repeatedly, stuff like they have “deeply considered these questions and implemented a new submission policy that we believe will ensure our content uses AI technology responsibly by creators and customers alike”, and that “generative AI is a major leap forward for creators, leveraging machine learning’s incredible power to ideate faster by developing imagery using words, sketches, and gestures”.

Creators? Fuck off! These people aren’t creating anything! They’re punching words into a computer that has been fed actual art! And even if Adobe can, as they’re claiming, only release images that have been “properly built, used, and disclosed”, it still sucks! Gah! Attempting to make good on one of AI art’s issues—art theft—doesn’t absolve it from its others, like the fact nothing to do with these images or their creation has anything to do with art!

Reaction among artists has of course been as wildly negative as any other AI art announcement over the past six months, with some criticising the company while others resort to more traditional cries: namely, that artists simply pirate PhotoShop instead of giving this company another cent.



Read original article here

Here’s What Google Covered During Its AI@ Event

Google AI Research Scientist Timnit Gebru speaks onstage during Day 3 of TechCrunch Disrupt SF 2018 at Moscone Center on September 7, 2018 in San Francisco, California.
Photo: Kimberly White (Getty Images)

A recurring theme at Google’s AI@ event was one of supposed reflection and cautious optimism. That sentiment rang particularly true on the topic of AI Ethics, an area Google’s struggled with in the past and one increasingly important in the emerging world of wild-generative AI. Though Google touted its own AI principles and Response AI teams for years, it has faced fierce blowback from critics, particularly after firing multiple high profile AI researchers.

Google Vice President of Engineering Research Marian Croak acknowledged some potential pitfalls presented by the technologies on display Wednesday. Those include fears around increased toxicity and bias heightened by algorithms, further degrading trust in news through deep fakes, and misinformation that can effectively blur the distinction between what is real and what isn’t. Part of that process, according to Croak, involves conducting research that creates the ability for users to have more control over AI systems so that they are collaborating with systems versus letting the system take full control of situations.

Croak said she believed Google’s AI Principles put users and the avoidance of harm and safety “above what our typical business considerations are.” Responsible AI researchers, according to Croak, conduct adversarial testing and set quantitative benchmarks across all dimensions of its AI. Researchers conducting those efforts are professionally diverse, and reportedly include social scientists, ethicists, and engineers among their mix.

“I don’t want the principles to just be words on paper,” Croak said. In the coming years, she said she hopes to see the capabilities of responsible AI embedded in the company’s technical infrastructure. Responsible AI, Croak said, should be “baked into the system.”

Read original article here

Microsoft unveils $4,299 Surface desktop computer



CNN
 — 

Microsoft’s most expensive Surface device is about to get even pricier.

At a press event on Wednesday, Microsoft is set to unveil several Surface Pro tablets, Surface Laptop models and a Surface Studio 2+ desktop computer, the last of which has not been updated in several years.

The new 28-inch Surface Studio 2+, an all-in-one desktop, now has an Intel Core H-35 processor, 50% faster CPU performance and an updated NVIDIA chip for faster graphics. The device also includes an updated display, cameras, microphones and supports a digital pen for on-screen drawing. It also has several ports, including USB with Thunderbolt 4, and the display can split into four different apps at once for greater multitasking.

The Surface Studio 2+ starts at $4,299, and $4,499 with the digital pen. The previous Surface Studio 2, released in 2018, received some criticism for its $3,499 starting price. Microsoft told CNN Business this year’s price jump is attributed to several significant improvements, including the new processor, a 1 TB SSD hard drive for faster file transfers and an enhanced 1080p camera, among other features.

The announcements about the refreshed Surface product lineup will kick off Microsoft’s days-long Ignite developer conference on Wednesday. The event comes as Microsoft marks the tenth anniversary of the Surface line, which originally launched with a tablet to take on the iPad.

Like other tech companies that have unveiled new products this fall, Microsoft is also confronting a more difficult economic environment, including high inflation and fears of a looming recession, that could make it harder to convince customers to spend three or even four figures upgrading devices.

While the new Surface products aren’t much different in terms of design or screen size than previous iterations, the latest devices feature some upgrades, including new chipsets for better performance.

Microsoft showed off its flagship Surface Pro 9 tablet, once again aimed at replacing the laptop. The two-in-one device features an aluminum casing in new colors as well as a built-in kickstand and a PixelSense display. Underneath the display is an HD camera, updated speakers and microphones, and a custom G6 chip. Microsoft said the chip helps power apps with digital ink, such as Ink Focus in Microsoft OneNote and the GoodNotes app for Windows 11, which is designed to make it feel like the user is writing with a pen and paper.

The Surface Pro 9 also offers a choice between processors. The first option is a 12th Gen Intel Core processor built on the Intel Evo platform 4 with Thunderbolt 4 – a combination which promises 50% more performance, better multitasking and desktop productivity, faster data transfer, and the ability to dock to multiple 4K displays. The second option is a Microsoft SQ3 processor powered by Qualcomm Snapdragon with 5G connectivity, with up to 19 hours of battery and new AI features.

The Surface Pro 9 is available in four colors, including platinum, graphite, sapphire and forest. It starts at $999.

Microsoft also introduced an update to its ultra-portable laptop, Surface Laptop 5, which looks very similar to its predecessor but with a processor update that may attempt to bring it closer in competition with Apple’s ARM-based chipsets for macOS laptops.

Surface Laptop 5 runs on Intel Evo platform and comes in two display sizes: 13.5 inches and 15 inches. It comes with updated Dolby Atmos 3D spatial speakers, a front-facing HD camera that automatically adjusts camera exposure in any lighting, and several new aluminum colors, such as cool metal, sage and alcantara. The company also said it promises one day of battery life on a single charge and is 50% more powerful than its predecessor.

The Surface Laptop starts at $999 for the 13.5-inch version and $1299 for the 15 inch. Pre-orders begin for Surface products on Wednesday in select markets and start hitting shelves later this month.

Microsoft hardware devices amount to between 3% to 5% of the tablet market, according to David McQueen, an analyst at ABI Research. Instead, the bulk of its revenue comes from Microsoft OS across different device types and associated applications and cloud services.

“Microsoft is able to stay in the hardware sector because of revenue generated from these services,” McQueen said. It’s an approach similar to Google whose Pixel smartphone remains a niche product but serves as a way for the company to highlight its apps and OS.

On Wednesday, the company also announced a new Microsoft Designer app and Image Creator in Bing and the Edge browser to bring advanced graphic design to mainstream audiences. The platform relies heavily on a partnership with startup OpenAI and its AI-powered DALL-E 2 tool, which generates custom images using text prompts. DALL-E 2 is also coming to Microsoft’s Azure OpenAI Service.

Brands are increasingly using DALL-E 2 for both ads and product inspiration, according to Microsoft. In a blog post, the company detailed how toy company Mattel sought out DALL-E 2 to conceptualize how future cars may look, such as by changing colors and typing “make it a convertible,” among other commands.

Experts in the AI field have raised concerns that the open-ended nature of these systems — which makes them adept at generating all kinds of images from words — and their ability to automate image-making means they could automate bias on a massive scale. In previous test of OpenAI’s system, for example, typing in “CEO” showed images that all appeared to be men and nearly all of them were white.

Microsoft said it is taking the concerns seriously. Inappropriate text requests will be denied by Microsoft’s servers, according to the company, and users will ultimately be banned for repeat offenses.

Read original article here

Silicon Valley’s next trillion-dollar companies?

Stable Diffusion’s web interface, DreamStudio

Screenshot/Stable Diffusion

Computer programs can now create never-before-seen images in seconds.

Feed one of these programs some words, and it will usually spit out a picture that actually matches the description, no matter how bizarre.

The pictures aren’t perfect. They often feature hands with extra fingers or digits that bend and curve unnaturally. Image generators have issues with text, coming up with nonsensical signs or making up their own alphabet.

But these image-generating programs — which look like toys today — could be the start of a big wave in technology. Technologists call them generative models, or generative AI.

“In the last three months, the words ‘generative AI’ went from, ‘no one even discussed this’ to the buzzword du jour,” said David Beisel, a venture capitalist at NextView Ventures.

In the past year, generative AI has gotten so much better that it’s inspired people to leave their jobs, start new companies and dream about a future where artificial intelligence could power a new generation of tech giants.

The field of artificial intelligence has been having a boom phase for the past half-decade or so, but most of those advancements have been related to making sense of existing data. AI models have quickly grown efficient enough to recognize whether there’s a cat in a photo you just took on your phone and reliable enough to power results from a Google search engine billions of times per day.

But generative AI models can produce something entirely new that wasn’t there before — in other words, they’re creating, not just analyzing.

“The impressive part, even for me, is that it’s able to compose new stuff,” said Boris Dayma, creator of the Craiyon generative AI. “It’s not just creating old images, it’s new things that can be completely different to what it’s seen before.”

Sequoia Capital — historically the most successful venture capital firm in the history of the industry, with early bets on companies like Apple and Google — says in a blog post on its website that “Generative AI has the potential to generate trillions of dollars of economic value.” The VC firm predicts that generative AI could change every industry that requires humans to create original work, from gaming to advertising to law.

In a twist, Sequoia also notes in the post that the message was partially written by GPT-3, a generative AI that produces text.

How generative AI works

Image generation uses techniques from a subset of machine learning called deep learning, which has driven most of the advancements in the field of artificial intelligence since a landmark 2012 paper about image classification ignited renewed interest in the technology.

Deep learning uses models trained on large sets of data until the program understands relationships in that data. Then the model can be used for applications, like identifying if a picture has a dog in it, or translating text.

Image generators work by turning this process on its head. Instead of translating from English to French, for example, they translate an English phrase into an image. They usually have two main parts, one that processes the initial phrase, and the second that turns that data into an image.

The first wave of generative AIs was based on an approach called GAN, which stands for generative adversarial networks. GANs were famously used in a tool that generates photos of people who don’t exist. Essentially, they work by having two AI models compete against each other to better create an image that fits with a goal.

Newer approaches generally use transformers, which were first described in a 2017 Google paper. It’s an emerging technique that can take advantage of bigger datasets that can cost millions of dollars to train.

The first image generator to gain a lot of attention was DALL-E, a program announced in 2021 by OpenAI, a well-funded startup in Silicon Valley. OpenAI released a more powerful version this year.

“With DALL-E 2, that’s really the moment when when sort of we crossed the uncanny valley,” said Christian Cantrell, a developer focusing on generative AI.

Another commonly used AI-based image generator is Craiyon, formerly known as Dall-E Mini, which is available on the web. Users can type in a phrase and see it illustrated in minutes in their browser.

Since launching in July 2021, it’s now generating about 10 million images a day, adding up to 1 billion images that have never existed before, according to Dayma. He’s made Craiyon his full-time job after usage skyrocketed earlier this year. He says he’s focused on using advertising to keep the website free to users because the site’s server costs are high.

A Twitter account dedicated to the weirdest and most creative images on Craiyon has over 1 million followers, and regularly serves up images of increasingly improbable or absurd scenes. For example: An Italian sink with a tap that dispenses marinara sauce or Minions fighting in the Vietnam War.

But the program that has inspired the most tinkering is Stable Diffusion, which was released to the public in August. The code for it is available on GitHub and can be run on computers, not just in the cloud or through a programming interface. That has inspired users to tweak the program’s code for their own purposes, or build on top of it.

For example, Stable Diffusion was integrated into Adobe Photoshop through a plug-in, allowing users to generate backgrounds and other parts of images that they can then directly manipulate inside the application using layers and other Photoshop tools, turning generative AI from something that produces finished images into a tool that can be used by professionals.

“I wanted to meet creative professionals where they were and I wanted to empower them to bring AI into their workflows, not blow up their workflows,” said Cantrell, developer of the plug-in.

Cantrell, who was a 20-year Adobe veteran before leaving his job this year to focus on generative AI, says the plug-in has been downloaded tens of thousands of times. Artists tell him they use it in myriad ways that he couldn’t have anticipated, such as animating Godzilla or creating pictures of Spider-Man in any pose the artist could imagine.

“Usually, you start from inspiration, right? You’re looking at mood boards, those kinds of things,” Cantrell said. “So my initial plan with the first version, let’s get past the blank canvas problem, you type in what you’re thinking, just describe what you’re thinking and then I’ll show you some stuff, right?”

An emerging art to working with generative AIs is how to frame the “prompt,” or string of words that lead to the image. A search engine called Lexica catalogs Stable Diffusion images and the exact string of words that can be used to generate them.

Guides have popped up on Reddit and Discord describing tricks that people have discovered to dial in the kind of picture they want.

Startups, cloud providers, and chip makers could thrive

Image generated by DALL-E with prompt: A cat on sitting on the moon, in the style of Pablo Picasso, detailed, stars

Screenshot/OpenAI

Some investors are looking at generative AI as a potentially transformative platform shift, like the smartphone or the early days of the web. These kinds of shifts greatly expand the total addressable market of people who might be able to use the technology, moving from a few dedicated nerds to business professionals — and eventually everyone else.

“It’s not as though AI hadn’t been around before this — and it wasn’t like we hadn’t had mobile before 2007,” said Beisel, the seed investor. “But it’s like this moment where it just kind of all comes together. That real people, like end-user consumers, can experiment and see something that’s different than it was before.”

Cantrell sees generative machine learning as akin to an even more foundational technology: the database. Originally pioneered by companies like Oracle in the 1970s as a way to store and organize discrete bits of information in clearly delineated rows and columns — think of an enormous Excel spreadsheet, databases have been re-envisioned to store every type of data for every conceivable type of computing application from the web to mobile.

“Machine learning is kind of like databases, where databases were a huge unlock for web apps. Almost every app you or I have ever used in our lives is on top of a database,” Cantrell said. “Nobody cares how the database works, they just know how to use it.”

Michael Dempsey, managing partner at Compound VC, says moments where technologies previously limited to labs break into the mainstream are “very rare” and attract a lot of attention from venture investors, who like to make bets on fields that could be huge. Still, he warns that this moment in generative AI might end up being a “curiosity phase” closer to the peak of a hype cycle. And companies founded during this era could fail because they don’t focus on specific uses that businesses or consumers would pay for.

Others in the field believe that startups pioneering these technologies today could eventually challenge the software giants that currently dominate the artificial intelligence space, including Google, Facebook parent Meta and Microsoft, paving the way for the next generation of tech giants.

“There’s going to be a bunch of trillion-dollar companies — a whole generation of startups who are going to build on this new way of doing technologies,” said Clement Delangue, the CEO of Hugging Face, a developer platform like GitHub that hosts pre-trained models, including those for Craiyon and Stable Diffusion. Its goal is to make AI technology easier for programmers to build on.

Some of these firms are already sporting significant investment.

Hugging Face was valued at $2 billion after raising money earlier this year from investors including Lux Capital and Sequoia; and OpenAI, the most prominent startup in the field, has received over $1 billion in funding from Microsoft and Khosla Ventures.

Meanwhile, Stability AI, the maker of Stable Diffusion, is in talks to raise venture funding at a valuation of as much as $1 billion, according to Forbes. A representative for Stability AI declined to comment.

Cloud providers like Amazon, Microsoft and Google could also benefit because generative AI can be very computationally intensive.

Meta and Google have hired some of the most prominent talent in the field in hopes that advances might be able to be integrated into company products. In September, Meta announced an AI program called “Make-A-Video” that takes the technology one step farther by generating videos, not just images.

“This is pretty amazing progress,” Meta CEO Mark Zuckerberg said in a post on his Facebook page. “It’s much harder to generate video than photos because beyond correctly generating each pixel, the system also has to predict how they’ll change over time.”

On Wednesday, Google matched Meta and announced and released code for a program called Phenaki that also does text to video, and can generate minutes of footage.

The boom could also bolster chipmakers like Nvidia, AMD and Intel, which make the kind of advanced graphics processors that are ideal for training and deploying AI models.

At a conference last week, Nvidia CEO Jensen Huang highlighted generative AI as a key use for the company’s newest chips, saying these kind of programs could soon “revolutionize communications.”

Profitable end uses for Generative AI are currently rare. A lot of today’s excitement revolves around free or low-cost experimentation. For example, some writers have been experimented with using image generators to make images for articles.

One example of Nvidia’s work is the use of a model to generate new 3D images of people, animals, vehicles or furniture that can populate a virtual game world.

Ethical issues

Prompt: “A cat sitting on the moon, in the style of picasso, detailed”

Screenshot/Craiyon

Ultimately, everyone developing generative AI will have to grapple with some of the ethical issues that come up from image generators.

First, there’s the jobs question. Even though many programs require a powerful graphics processor, computer-generated content is still going to be far less expensive than the work of a professional illustrator, which can cost hundreds of dollars per hour.

That could spell trouble for artists, video producers and other people whose job it is to generate creative work. For example, a person whose job is choosing images for a pitch deck or creating marketing materials could be replaced by a computer program very shortly.

“It turns out, machine-learning models are probably going to start being orders of magnitude better and faster and cheaper than that person,” said Compound VC’s Dempsey.

There are also complicated questions around originality and ownership.

Generative AIs are trained on huge amounts of images, and it’s still being debated in the field and in courts whether the creators of the original images have any copyright claims on images generated to be in the original creator’s style.

One artist won an art competition in Colorado using an image largely created by a generative AI called MidJourney, although he said in interviews after he won that he processed the image after choosing it from one of hundreds he generated and then tweaking it in Photoshop.

Some images generated by Stable Diffusion seem to have watermarks, suggesting that a part of the original datasets were copyrighted. Some prompt guides recommend using specific living artists’ names in prompts in order to get better results that mimic the style of that artist.

Last month, Getty Images banned users from uploading generative AI images into its stock image database, because it was concerned about legal challenges around copyright.

Image generators can also be used to create new images of trademarked characters or objects, such as the Minions, Marvel characters or the throne from Game of Thrones.

As image-generating software gets better, it also has the potential to be able to fool users into believing false information or to display images or videos of events that never happened.

Developers also have to grapple with the possibility that models trained on large amounts of data may have biases related to gender, race or culture included in the data, which can lead to the model displaying that bias in its output. For its part, Hugging Face, the model-sharing website, publishes materials such as an ethics newsletter and holds talks about responsible development in the AI field.

“What we’re seeing with these models is one of the short-term and existing challenges is that because they’re probabilistic models, trained on large datasets, they tend to encode a lot of biases,” Delangue said, offering an example of a generative AI drawing a picture of a “software engineer” as a white man.



Read original article here

Elon Musk Unveils Prototype of Tesla’s Humanoid Robot Optimus, Says It Will Cost Less Than a Car

Mr. Musk first laid out the vision for the robot, called Optimus, a little more than a year ago at Tesla’s first-ever AI day. At the time, a dancer in a costume appeared onstage. This time, Mr. Musk presented a prototype at the gathering that unfolded late Friday in Palo Alto, Calif.

The early prototype, which still had wires showing, took a few steps, waved to the crowd, and performed some basic dance moves.

Tesla’s robot is expected to cost less than a car, with a price point below $20,000, Elon Musk said.



Photo:

Tesla

Mr. Musk quipped the robot could do a lot more, but limited its activity for fear it could fall on its face. The robot’s appearance on stage marked the first time it operated without a tether, Mr. Musk said.

“Our goal is to make a useful humanoid robot as quickly as possible,” he said, with the aspiration of being able to make them at high volume and low cost. “It is expected to cost much less than a car,” he said, with a price point below $20,000. Customers should be able to receive the robot, once ordered, in three to five years, Mr. Musk said. It isn’t yet for sale.

He later showed off a nonfunctioning, sleeker model that he said was closer to the production version.

“There’s still a lot of work to be done to refine Optimus,” he said, saying that the concept could evolve over time. “It won’t be boring.”

The battery-powered robot should be able to handle difficult chores, Tesla said, including lifting a half-ton, 9-foot concert grand piano. Mr. Musk added it would have conversational capabilities and feature safeguards to prevent wrongdoing by the machine.

Elon Musk last year unveiled the idea of the robot Optimus with a dancer in a costume.



Photo:

TESLA/via REUTERS

“I’m a big believer in AI safety,” said Mr. Musk, who has previously expressed concerns about how such technology could be used. He said he thinks there should be a regulatory authority at the government level.

The Tesla boss painted a vision of Optimus as helping Tesla make cars more efficiently, starting with simple tasks and then expanded uses. He has also suggested the robot could serve broader functions and potentially alleviate labor shortages.

“It will, I think, turn the whole notion of what’s an economy on its head, at the point at which you have no shortage of labor,” Mr. Musk said Aug. 4 at Tesla’s annual shareholder meeting. On Friday, he added: “It really is a fundamental transformation of civilization as we know it.”

Elon Musk unveiled a prototype of Tesla’s humanoid robot Optimus, part of an effort to shape perception of the company as more than just a car maker. The Tesla CEO said the robot is expected to cost less than a car. Photo: Tesla

When he first unveiled the Optimus concept, Mr. Musk said such a robot could have such an impact on the labor market it could make it necessary to provide a universal basic income, or a stipend to people without strings attached.

Tesla has also encountered problems with automation. Early efforts to rely heavily on automated tools to scale up vehicle production suffered setbacks, and the company had to rely more heavily than planned on factory workers. Mr. Musk later tweeted: “Yes, excessive automation at Tesla was a mistake. To be precise, my mistake. Humans are underrated.”

One of the big questions around Tesla’s humanoid robot is its central purpose, said

Chris Atkeson,

a Carnegie Mellon University robotics professor. If Tesla’s main goal is to improve manufacturing, a quadruped likely would have been easier to build than a humanoid robot, in part because additional legs make it easier to balance, he said.

SHARE YOUR THOUGHTS

What do you think of the Tesla robot? Join the conversation below.

Mr. Musk, who has been instrumental in popularizing electric vehicles and pioneered landing rocket boosters with his company SpaceX, also has a record of making bold predictions that don’t immediately pan out. Three years ago at an event about automation, he projected that more than a million Tesla vehicles would be able to operate without a driver by the middle of 2020, positioning the company to launch a robot taxi service. That hasn’t happened.

Mr. Musk for some time has said Tesla aimed to be more than just a car company and reiterated that message on Friday. He called the company “a series of startups.”

Mr. Musk billed the latest event, like last year’s, as one aimed at recruiting engineers in fields such as artificial intelligence, robotics and chips.

Tesla has long bet on automation to keep the company ahead of competitors. The company’s cars are outfitted with an advanced driver-assistance system, known as Autopilot, that helps drivers with tasks such as maintaining a safe distance from other vehicles on the road and staying centered in a lane.

Tesla engineers detailed some of the AI work the company is doing, including to underpin its driver-assistance technology. Mr. Musk said the company’s development of a powerful, AI-focused computer could allow Tesla to offer the number-crunching capability as a service to others, not unlike cloud-computing offerings provided by the likes of

Amazon.com Inc.

The company is developing and selling an enhanced version of Autopilot that brings more automated driving into cities. Tesla calls the system Full Self-Driving, or FSD, although it doesn’t actually make vehicles autonomous and the company tells drivers to keep their hands on the wheel while operating the car.

Tesla said Friday that it now has 160,000 customers with the software. Mr. Musk said rollout of the technology beyond the U.S. and Canada depends on gaining regulatory approval, though it should be feasible from a technology perspective by year-end.

Tesla has steadily raised the price of FSD, which now retails for $15,000. AI has been at the heart of Tesla’s efforts to develop more advanced driver-assistance features and, eventually, fully autonomous vehicles.

Tesla said the software that is used to take on more driving functions also underpins operations of the humanoid robot.

Tesla’s pursuit of automation has increasingly come under scrutiny. The National Highway Traffic Safety Administration, which regulates auto safety, opened a probe into Autopilot last year after a series of crashes involving Teslas that struck first-responder vehicles stopped for roadway emergencies.

Two U.S. senators have also asked the Federal Trade Commission to investigate whether Tesla has been deceptive in its marketing of Autopilot and FSD.

The electric-car maker has long said that driving with Autopilot engaged is safer than doing so without it. Tesla points to internal data showing that crashes were less common when drivers were using Autopilot, though some researchers have criticized the company’s methodology.

Write to Meghan Bobrowsky at Meghan.Bobrowsky@wsj.com and Rebecca Elliott at rebecca.elliott@wsj.com

Copyright ©2022 Dow Jones & Company, Inc. All Rights Reserved. 87990cbe856818d5eddac44c7b1cdeb8

Read original article here

Tesla robot slowly walks on stage at AI Day


Washington, DC
CNN
 — 

Tesla revealed on Friday a prototype of a humanoid robot that it says could be a future product for the automaker.

The robot, dubbed Optimus by Tesla, walked stiffly on stage at Tesla’s AI Day, slowly waved at the crowed and gestured with its hands for roughly one minute. Tesla CEO Elon Musk said that the robot was operating without a tether for the first time. Robotics developers often use tethers to support robots because they aren’t capable enough to walk without falling and damaging themselves.

The Optimus’ abilities appear to significantly trail what robots from competitors like Hyundai-owned Boston Dynamics are capable of. Boston Dynamics robots have been seen doing back flips and performing sophisticated dance routines without a tether.

“The robot can actually do a lot more than we just showed you,” Musk said at the event. “We just didn’t want it to fall on its face.”

Tesla also showed videos of its robot performing simple tasks like carrying boxes and watering plants with a watering can.

Musk claimed that if the robot was produced in mass volumes it would “probably” cost less than $20,000. Tesla maintains that Optimus’ advantage over competitors will be its ability to navigate independently using technology developed from Tesla’s driver-assistance system “Full Self Driving,” as well as cost savings from what it has learned about manufacturing from its automotive division. (Tesla’s “Full Self Driving” requires a human that is alert and attentive, ready to take over at any time, as it is not yet capable of fully driving itself.)

Tesla has a history of aggressive price targets that it doesn’t ultimately reach. The Tesla Model 3 was long promised as a $35,000 vehicle, but could only very briefly be purchased for that price, and not directly on its website. The most affordable Tesla Model 3 now costs $46,990. When Tesla revealed the Cybertruck in 2019, its pick-up truck that remains unavailable for purchase today, it was said to cost $39,990, but the price has since been removed from Tesla’s website.

Tesla AI Day is intended largely as a recruiting event to attract talented people to join the company.

Musk claimed the robot could be transformative for civilization. The robot displayed Friday, despite its limitations compared to competitors, was significantly ahead of what Tesla revealed a year ago, when a person jumped on stage in a robot suit and danced around.

“‘Last year was just a person in a robot suit,” Musk said before the robot walked on stage. “We’ve come a long way. Compared to that, it’s going to be very impressive.”

Tesla is not the first automaker to develop a humanoid robot. Along with Hyundai’s Boston Dynamics, Honda worked on robots dubbed “Asimo” for nearly 20 years. In its final form, Asimo was a child-size humanoid robot capable of untethered walking, running, climbing and descending stairs, and manipulating objects with its fingers.

Read original article here