AI-Controlled VTuber Streams Games On Twitch, Denies Holocaust

Screenshot: Vedal / Twitch / Kotaku

Neuro-sama is a VTuber who streams Minecraft and the rhythm game Osu! on Twitch. But unlike most anime avatars, she’s controlled by an artificial intelligence program rather than a human being. That makes her catnip for the denizens of Twitch chat, who can prompt her to respond with all sorts of questions ranging from innocent inquiries to 4chan trolling. Within the first few streams, someone had already asked Neuro-sama about the Holocaust. “I’m not sure if I believe it,” she said.

That was one of the more infamous clips that went viral online near the end of last month. Asked what she thought of women’s rights, she said they didn’t exist. How would she solve philosophy’s famous trolley ethical conundrum? Throw a fat person on the tracks. Often, however, she’ll go for long stretches without getting tempted by the chat into controversial or hateful remarks. In that way she’s an impressive simulacrum of a Twitch streamer straddling the chasm between repetitive banter and edgelord antics.

“The controversial things she says is due to the fact that she tries to make witty and comical remarks about whatever is said in chat, aligning AIs with human values is an ongoing area of research,” Neuro-sama’s creator, a game programmer named Vedal, told Kotaku. “To counter this, I’ve worked hard since the first few streams to improve the strength of the filters used for her. Data that she learns on is also manually curated to mitigate negative biases. We now also have a team of people moderating twitch chat who check everything she says.”

Neuro-sama isn’t Vedal’s first AI. In fact, a version of her was first created years ago with the explicit purpose of learning to play Osu!, a long running free-to-play rhythm game where you click shapes on a screen to the beat of anime music. While those sessions were also streamed, there was no avatar or interactive personality. Following last year’s surge of big-name VTubers, Neuro-sama builds on the Osu! skills of the original project with a fully-voiced Twitch performance that can riff with the audience.

It’s perfect timing given the internet’s recent love affair with the OpenAI-powered ChatGPT chatbot, where users could submit hyper-specific text prompts and receive uncannily artful responses in return. Vedal wouldn’t go into detail about how Neuro-sama learns and communicates, other than to confirm she relies on a large language model, which has been “trained on a large amount of text on the internet.” While not as sophisticated, the effect has been convincing enough to net Neuro-sama thousands of viewers per stream.

She also recently defeated the top-ranked Osu! player, Mrekk, on December 28, though some fans of the game debate whether the human opponent was ill-served by the song selection. Neuro-sama has since moved onto Minecraft, a much more complex game with far more possibilities for unexpected moments as players ask whether Melee is the best Super Smash Bros. and whether she’ll step on them. The moderation tools are apparently better now too.

“She picks what to respond to within a limited window,” Vedal said. “However it should be noted that she will not talk about the Holocaust as the filters have been improved.” Instead, she’s currently trying to learn how to sing.

via Kotaku

January 6, 2023 at 03:25PM

Welcome to the Wet Hot AI Chatbot Summer

Late last year, I attended an event hosted by Google to celebrate its AI advances. The company’s domain in New York’s Chelsea neighborhood now extends literally onto the Hudson River, and about a hundred of us gathered in a pierside exhibition space to watch scripted presentations from executives and demos of the latest advances. Speaking remotely from the West Coast, the company’s high priest of computation, Jeff Dean, promised “a hopeful vision for the future.”  

The theme of the day was “exploring the (im)possible.” We learned how Google’s AI was being put to use fighting wildfires, forecasting floods, and assessing retinal disease. But the stars of this show were what Google called “generative AI models.” These are the content machines, schooled on massive training sets of data, designed to churn out writings, images, and even computer code that once only humans could hope to produce.

Something weird is happening in the world of AI. In the early part of this century, the field burst out of a lethargy—known as an AI winter—by the innovation of “deep learning” led by three academics. This approach to AI transformed the field and made many of our applications more useful, powering language translations, search, Uber routing, and just about everything that has “smart” as part of its name. We’ve spent a dozen years in this AI springtime. But in the past year or so there has been a dramatic aftershock to that earthquake as a sudden profusion of mind-bending generative models have appeared.

Most of the toys Google demoed on the pier in New York showed the fruits of generative models like its flagship large language model, called LaMDA. It can answer questions and work with creative writers to make stories. Other projects can produce 3D images from text prompts or even help to produce videos by cranking out storyboard-like suggestions on a scene-by-scene basis. But a big piece of the program dealt with some of the ethical issues and potential dangers of unleashing robot content generators on the world. The company took pains to emphasize how it was proceeding cautiously in employing its powerful creations. The most telling statement came from Douglas Eck, a principal scientist at Google Research. “Generative AI models are powerful—there’s no doubt about that,” he said. “But we also have to acknowledge the real risks that this technology can pose if we don’t take care, which is why we’ve been slow to release them. And I’m proud we’ve been slow to release them.”  

But Google’s competitors don’t seem to have “slow” in their vocabularies. While Google has provided limited access to LaMDA in a protected Test Kitchen app, other companies have been offering an all-you-can-eat smorgasbord with their own chatbots and image generators. Only a few weeks after the Google event came the most consequential release yet: OpenAI’s latest version of its own powerful text generation technology, ChatGPT, a lightning-fast, logorrheic gadfly that spits out coherent essays, poems, plays, songs, and even obituaries at the merest hint of a prompt. Taking advantage of the chatbot’s wide availability, millions of people have tinkered with it and shared its amazing responses, to the point where it’s become an international obsession, as well as a source of wonder and fear. Will ChatGPT kill the college essay? Destroy traditional internet search? Put millions of copywriters, journalists, artists, songwriters, and legal assistants out of a job?

Answers to those questions aren’t clear right now. But one thing is. Granting open access to these models has kicked off a wet hot AI summer that’s energizing the tech sector, even as the current giants are laying off chunks of their workforces. Contrary to Mark Zuckerberg’s belief, the next big paradigm isn’t the metaverse—it’s this new wave of AI content engines, and it’s here now. In the 1980s, we saw a gold rush of products moving tasks from paper to PC application. In the 1990s, you could make a quick fortune by shifting those desktop products to online. A decade later, the movement was to mobile. In the 2020s the big shift is toward building with generative AI. This year thousands of startups will emerge with business plans based on tapping into the APIs of those systems. The cost of churning out generic copy will go to zero. By the end of the decade, AI video-generation systems may well dominate TikTok and other apps. They may not be anywhere as good as the innovative creations of talented human beings, but the robots will quantitatively dominate.

via Wired Top Stories

January 6, 2023 at 08:12AM

South Korea’s Lunar Orbiter Captures Unreal Views of Our Home Planet

The Danuri Lunar Orbiter caught this view of Earth.
Image: KARI

From its position in low lunar orbit, South Korea’s first Moon mission caught a unique glimpse of Earth rising from behind the cratered surface of our natural satellite.

The Korean Pathfinder Lunar Orbiter (KPLO), also known as Danuri, beamed back beautiful black-and-white images of Earth captured by its high-resolution camera. The two images were taken on December 24 and 28 and released by the Korean Aerospace Research Institute on Monday.

The images show a solemn Earth in the distance, while the Moon’s dusty surface appears in the foreground. From our perspective on Earth, we often see the Moon rising above our planet’s surface. But the images taken by Danuri from lunar orbit provide a counterintuitive view of our home planet glimmering behind the Moon’s surface.

Danuri captured the first image when it was 77 miles (124 kilometers) above the lunar surface and the second when it was around 213 miles (344 kilometers) above its surface.

G/O Media may get a commission

Danuri launched on August 5 on board SpaceX’s Falcon 9 rocket, marking South Korea’s first deep space mission. On December 17, the spacecraft completed its first lunar orbit insertion maneuver and entered into lunar orbit.

The 1,100-pound (500-kilogram) probe is equipped with four science instruments built locally, as well as a NASA camera to capture views of the lunar surface. From its low orbit, Danuri will explore the Moon’s shadowed regions, which could hold water ice.

By launching Danuri, South Korea hopes to advance its lunar exploration, as the orbiter is designed to scope out potential landing spots for future missions to the Moon. South Korea also wants to launch a lander and a rover, in addition to another orbiter, for the second phase of the mission.

More: Chinese Mission to Pluck Samples from Moon’s Far Side Just Got More Interesting

via Gizmodo

January 5, 2023 at 03:57PM

Valve Could Allow Local Network Downloads For Steam Deck

It looks like Valve might enable local network downloads for Steam Deck, making it easier and quicker to get games on the handheld.

Pavel Djundik, creator of Steam DB, posted on Twitter that Valve had added more code potentially allowing users to move files from their PC to their Steam Deck through local networks. Djundik previously tweeted about Steam working on peer-to-peer downloads in October 2022.

Now Playing: Steam Deck – Everything To Know

As some Twitter users pointed out, it’s like Warpinator–where users can send and transfer files over local networks–but for the Steam ecosystem.

For now, Steam Deck downloads still need to be done through wi-fi, which for particularly large games, can take a long time. Or if you have more than one gamer in the house, the process takes even longer.

The Steam Deck is still relatively new, but there’s definitely a second-generation Steam Deck planned. According to the designers who spoke to The Verge in an interview, they’d like to focus on improving the console’s battery life and screen in the next iteration. It’s unknown when we can expect to see the future V2 Steam Deck.

In other Steam news, Valve released the results of its 2022 Steam Awards. Elden Ring nabbed both Game of the Year and Best Game You Suck At. Outstanding Story-Rich Game went to God of War and Best Game On The Go went to Death Stranding Director’s Cut.

Got a news tip or want to contact us directly? Email

via GameSpot’s PC Reviews

January 6, 2023 at 12:02PM