However, most people didn't expect one of those features was incorporating OpenAI's ChatGPT technology into the game. Supernova uses ChatGPT to allow players to create their own alien civilizations with just some text prompts.

We got in touch with Wardell by email to ask him some more questions about Galactic Civilizations IV: Supernova, and how he thinks ChatGPT will cause Stardock to actually bring in more team members to develop games.

The Galactic Civilizations series has been ongoing for 30 years now. What are your thoughts about continuing to make games in this franchise for such a long time?

Brad Wardell - I’ve long felt that as long as there were new and interesting things to try out in the 4X genre, I’d keep at it. The impetus for me, 30 years ago, was to try and see if you could make a multithreaded AI. An AI that would calculate in the background. Now, here we are, using a large language model like ChatGPT to give players unique, personalized experiences.

Do you believe the series has served as an inspiration for other 4x games over the decades?

Brad Wardell - Certainly. I’ve borrowed heavily from other games myself, not the least being the Civilization series. It’s always validating when a concept we came up with shows up in other games.

When it came to making the Supernova standalone expansion what were the dev team's main goals?

Brad Wardell - I would have preferred to make it a normal expansion. But, as we got in there, we realized we would need two copies of almost every data file to do that.

What really exploded the scale of this “expansion” has been the rise of AI. What started out as a fairly normal expansion process became something different when we realized we could use AI to create a real-time, living universe for players.

The integration with ChatGPT is one of the first such additions to a major PC game. How did the idea come about to add those features in Supernova?

Brad Wardell - Stardock has been working on multi-parameter AI for the past several years in another title we’re developing. So, we’ve been keeping up to date with OpenAI’s work for some time. As soon as they had an API we could use, we realized there was an opportunity to vastly increase the amount of content available for players and create a lot of new, emergent gameplay.

How hard or easy was it to incorporate the text prompts for ChatGPT so it could make new civilizations and quests for the game?

Brad Wardell - It took quite a bit of time. For one thing, we didn’t put it directly into the game. It’s hosted on our cloud services tech called Tachyon. This way, any of our games can make API calls to Tachyon and then it can use whatever is the best tech to deliver what is needed.

On the art side, we also had to make sure that the art was being trained on Stardock art and not, say, art from ArtStation, as we have a lot of artists here who have strong feelings about where the training data comes from.

Another challenge is making sure that the GPT results weren’t purely cosmetic. We want to be able to generate in-game quests and have it affect relationships. We want it to affect even what the ships look like, so that took quite a bit of time as well.

What are some of the more surprising things you have learned or encountered about adding ChatGPT into Supernova?

Brad Wardell - How well it works. We didn’t really expect the writing results to be as good as they’ve been. It helps that ChatGPT is very familiar with GalCiv and its style of writing already so that results tend to fall into our existing, slightly humorous, writing style right away.

In a forum post, you stated that adding ChatGPT into Supernova is actually causing Stardock to hire more people, not less. Can you go into specifics on that?

Brad Wardell - Indeed. It’s changing the types of people we’d be hiring. But let me walk you through the business reason for this, as I think many people haven’t really thought through the issue to its logical conclusion.

In game development, you need to have a critical mass of X in order for it to be worthwhile. People who think AI generated art or AI generated writing is going to eliminate all the jobs haven’t taken the critical mass feature threshold into account. AI will displace jobs, but that’s a different discussion.

For example, in the past, we just couldn’t justify many features or classes of content because even if we had 3 artists, they could only produce, say, 3 units of content. Therefore, we wouldn’t hire the artists. But now, those 3 artists can produce 300 units of content via AI assist. That amount of content exceeds the threshold needed for a feature to be worthwhile to have in the game and thus, 3 new art positions are created.

The same is true with writing. There’s now enough written content in the game to justify contracting a dedicated editor (and God knows, we’ve needed text editing for years). But, you need a lot of writing before you can make the business case of needing to hire writers and editors. The AI has allowed us to reach that threshold.

What other features do you consider to be important in Supernova?

Brad Wardell - A big part of it is the accessibility. We want to make GalCiv IV: Supernova have a lot more depth than GalCiv III (or even GalCiv IV base game on Epic).

And not to harp on the AI stuff again but as a practical matter, without AI assist tech like GitHub Copilot and our own internal GPT servers, a lot of these features would just be well beyond the scope of our team.

So, for example, Supernova has a new combat system, new planet management, cultural progression, minor raced system, star system upgrades, and core world upgrades, all with a UI designed to make it accessible for players new to the genre.

How will the game evolve during the Early Access period and when do you think it will be available for full access?

Brad Wardell - We are targeting late Summer or early Fall for the full release. We have a lot of features we are still looking to get into it. Our measure of success is if GalCiv IV: Supernova becomes the go-to game for turn-based 4X players.

What can you tell us about Stardock's current and upcoming game lineup, and is it possible those games will add ChatGPT features as well?

Brad Wardell - We have some unannounced games that make use of our multiparameter AI tech. I can’t say more than that right now.

Finally, is there anything else you want to say about Galactic Civilizations IV: Supernova?

Brad Wardell - Galactic Civilizations IV: Supernova represents a big step forward for the franchise. We can’t wait for people to try it out and for fans to see that we’ve been listening and writing down their ideas for the past several years since we released GalCiv III. The last GalCiv game I got to be this involved with was GalCiv II: Twilight of the Arnor, so on a personal note, this is a big treat for me.

We want to thank Brad Wardell for taking time to answer our questions. The game is available in Early Access for Steam and the Epic Games Store for $39.99. If you have already bought the original Galactic Civilizations IV on the Epic Games Store, you can get Supernova for a big discount.

Disclaimer: Neowin's relationship to Stardock

We interview Stardock's Brad Wardell about Galactic Civilizations IV: Supernova and ChatGPT

Conducted by researchers from Northeastern University, the study utilizes data from the US Department of Energy Manufacturing Energy Consumption Survey to assess the potential electricity production of rooftop solar systems in comparison to the electricity demand per unit of floor space for the average manufacturing facility. The findings reveal that, depending on the season, rooftop solar installations could fully meet the electricity needs of 5-35% of American manufacturing industries, with the furniture, textile, and apparel sectors standing to gain the most.

Dr. Matthew Eckelman, Associate Professor of Civil and Environmental Engineering at Northeastern University, says: “Currently, less than 0.1% of the electricity required by the manufacturing sector in the US is generated through renewable, on-site sources. This must change if we are going to meet decarbonization goals, and in many cases, rooftop solar panels are now a feasible option for supplying low-carbon energy.”

Globally, the industrial sector represents a large contributor to energy usage, and associated greenhouse gas and carbon emissions. As such, manufacturing has become an important target for global decarbonization efforts, with many companies switching to lower-carbon energy sources. The new study shows that rooftop solar panels could now be a feasible option for many manufacturing units due to their large, flat rooftops alongside falling prices, improved efficiencies, and flexibility in installation. Seasonally, manufacturing companies across nearly 40% of US locations could fulfill their electricity needs in the spring and summertime with rooftop solar arrays.

Eckelman concludes, “Greater policy attention on the feasibility and potential benefits of rooftop solar panel arrays will help industries to achieve renewable energy and greenhouse gas emissions goals. Our research provides an indication of the locations and sectors for which rooftop solar arrays could significantly help manufacturing firms to reach these goals.”

Source

Similarly, there is the way the private space sector has become a major participant, providing everything from launch services and commercial payloads to satellite constellations and crews.

But in some ways, old habits die hard. Since the turn of the century, China has emerged as a major power in space, to the point of becoming a direct competitor with NASA's human space programs. For the past few years, China has been developing a reusable autonomous spaceplane to compete with the X-37B Orbital Test Vehicle (OTV).

Known as Shenlong ("divine dragon"), this spaceplane recently concluded its second test flight after spending 276 days in orbit. Though the details are scant, the Chinese state media company Xinua declared the flight a breakthrough for the Chinese space program.

The Chinese reusable experimental spaceplane, the Chongfu Shiyong Shiyan Hangtian Qi (CSSHQ), has remained steeped in mystery since it first came to light. During its previous test flight (CSSHQ 1), the spaceplane took off on September 4, 2020, and spent a brief two days in orbit. For its second flight (CSSHQ-2), the Chinese spaceplane took off from the Jiuquan Satellite Launch Center in the Gobi Desert on August 4, 2022, atop a Long March-2F/T rocket (CZ-2) and reportedly landed at Juiquan again on Monday, May 8.

The Chinese state news agency Xinhua issued a statement shortly after the spaceplane landed (reportedly in Jiuquan). The spaceplane's manufacturer, the China Aerospace Science and Technology Corp. (CASC), reshared the Xinhua statement via the Chinese social media platform Weixin:

"The reusable test spacecraft successfully launched by our country at the Jiuquan Satellite Launch Center successfully returned to the scheduled landing site on May 8 after flying in orbit for 276 days. The complete success of this test marks an important breakthrough in my country's research on reusable spacecraft technology, which will provide a more convenient and inexpensive way to and from the peaceful use of space in the future."

No details were given about the mission, the spacecraft, or what operations were conducted in orbit. However, SpaceNews correspondent Andrew Jones recapped the CCSHQ's activities throughout the test flight and recapped them in a recent article.

"The spacecraft performed numerous small and much larger orbital maneuvers during its flight, with adjustments in recent weeks made in preparation for landing," he wrote. "The landing is likely to have taken place at the Lop Nur military base in Xinjiang. Information on the spacecraft's orbit suggests an orbital track over the facility around 0020 UTC provided the opportunity for landing."

As Jones wrote back in August 2022, the spacecraft also released a small satellite into orbit 90 days into its flight. While the purpose and nature of this satellite are unknown, tracking data obtained by the U.S. Space Force (USSF) revealed that the small satellite remained in very close proximity to the spaceplane. While this flight was a big step for China's research into reusable spacecraft technology, it pales in comparison to the accomplishments of the X-37B, which has executed six test flights since April 2010.

During its last flight (OTV-6), the OTV spent over nine-hundred days in space and even deployed a series of science experiments while in orbit. Like the CCSHQ's activities, the details about the OTV's flights have remained highly classified. However, various sources have confirmed that the objectives include developing orbital reconnaissance vehicles for military use and testing reusable spacecraft technologies, hypersonic engines, and autonomous guidance systems.

Concerning reusability, China is developing spacecraft as part of a larger program, including a super-heavy launch system similar to SpaceX's Starship.

The concept was presented by the China Academy of Launch Vehicle Technology (CALT) in April 2021 during the sixth annual Aerospace Industry Achievement Exhibition (aka. "National Space Day") in Nanjing.

The China Aerospace Science and Industry Corporation (CASIC) is also pursuing its own reusable spaceplane, known as the Tengyun ("cloud rider"). According to a video released by CASIC during the 2021 Zhuhai Air Show (shown above), this vehicle would be part of a two-stage launch system that relies on an air-launch vehicle (similar to Virgin Galactic).

This latest development showcases the progress China has made in recent years. In terms of space station technology, robotic exploration on the Moon (the Chang'e program) and Mars (the Tianwen-1 orbiter and Zhurong rover), and crewed spaceflight, China has emerged as a global powerhouse.

Later this decade, China plans to send the first taikonauts to the Moon and establish the International Lunar Research Station (ILRS), in direct competition with NASA's Artemis Program. In 2033, they hope to send the first crewed missions to Mars, precisely when NASA is hoping to do the same.

Source

However, that's not all that was added, according to the latest update to the official Bing blog.

One of them is adding images to some chat answers. Microsoft says:

For example, if you ask Bing chat questions about flamingos or capybaras, an image of these animals will appear in-line with your answer. Click on the image to launch a knowledge card that allows you to explore more information about their habitat, diet, lifecycle, and other facts.

One user on Twitter this week showed an image of a Bing Chat answer about the weather with a image that was taken from Microsoft's MSN Weather Service.

Another update this week was adding more optimized answers to Bing Chat turns. Microsoft stated:

We’ve redesigned the visual elements that appear at the end of text-based answers so they’re better designed for the Bing chat experience. This has been done across a variety of topics, with the intention of giving you more complete answers to your questions. For example: When using chat for shopping, we’ve made it much easier for you to compare side by side the items you’re considering.

Some more additions to Bing Chat this week include:

• Export: To help you put your chat answers to work, you can also now export them to a PDF, text file, or Microsoft Word document.
• Better copy and paste experience: We’ve also improved the copy and paste experience when Bing chat generates code or other blocks of formatted text. You’ll see a separate copy button that allows you to quickly drop it elsewhere.
• Prompts include text formatting: When writing or copying your prompt into Bing chat, you can now include formatting like paragraphs, bullets, or numbering.

Microsoft will almost certainly reveal more Bing Chat plans at its annual Build 2023 conference starting May 23.

Bing Chat adds visual answers for various topics and more in this week's update

When surgeons are trained, they usually need the supervision of more experienced doctors who can mentor them on their technique. That may be changing due to a new artificial intelligence system developed by Caltech researchers and Keck Medicine of USC urologists that aims to provide valuable feedback to surgeons on the quality of their work.

The goal of the new Surgical AI System (SAIS) is to provide surgeons with objective performance evaluations that can improve their work and, by extension, the outcomes of their patients. When provided with a video of a surgical procedure, SAIS can identify what type of surgery is being performed and the quality with which it was executed by a surgeon.

The system was introduced through a series of articles in the journals Nature Biomedical Engineering, npj Digital Medicine, and Communications Medicine, which were published concurrently at the end of March 2023.

“In high stakes environments such as robotic surgery, it is not realistic for AI to replace human surgeons in the short term,” says Anima Anandkumar, Bren Professor of Computing and Mathematical Sciences and senior author of the studies. “Instead, we asked how AI can safely improve surgical outcomes for the patients, and hence, our focus on making human surgeons better and more effective through AI.”

SAIS was trained using a large volume of video data that was annotated by medical professionals. Surgeons’ performances were assessed down to the level of individual discrete motions, i.e., holding a needle, driving it through tissue, and withdrawing it from tissue. After training, SAIS was tasked with reviewing and evaluating surgeons’ performance during a wide range of procedures using video from a variety of hospitals.

“SAIS has the potential to provide surgeon feedback that is accurate, consistent, and scalable,” says Dani Kiyasseh, lead author of the studies, a former postdoctoral researcher at Caltech and now a senior AI engineer at Vicarious Surgical. The hope, according to the researchers, is for SAIS to provide surgeons with guidance on what skill sets need to be improved.

To make the tool more valuable for surgeons, the team developed the AI’s ability to justify its skill assessments. The AI can now inform surgeons about their level of skill and provide detailed feedback on its rationale for making that assessment by pointing to specific video clips.

“We were able to show that such AI-based explanations often align with explanations that surgeons would have otherwise provided,” Kiyasseh says. “Reliable AI-based explanations can pave the way for providing feedback when peer surgeons are not immediately available.”

Early on, researchers testing SAIS noted that an unintended bias crept into the system in which the AI sometimes rated surgeons as more or less skilled than their experience would otherwise indicate based solely on an analysis of their overall movements. To address this issue, the researchers guided the AI system to focus exclusively on pertinent aspects of the surgical video. Narrowing the focus mitigated, though did not eliminate, the bias, which the researchers are continuing to address.

“Human-derived surgical feedback is not presently objective nor scalable,” says Andrew Hung, a urologist with Keck Medicine of USC and associate professor of urology at Keck School of Medicine of USC. “AI-derived feedback, such as what our system delivers, presents a major opportunity to provide surgeons actionable feedback.”

The studies are titled “A vision transformer for decoding surgeon activity from surgical videos,” “Human visual explanations mitigate bias in AI-based assessment of surgeon skills,” and “A multi-institutional study using artificial intelligence to provide reliable and fair feedback to surgeons.” This research was funded by the National Cancer Institute.

Source

Immediately one would notice that the Pulse RX 7600 is a 2-slot card, and it makes sense that cooler isn't over-engineered seeing how it will feature a single 8-pin power connector, similar to the RX 6600 Pulse. The packaging of the GPU is also pretty standard. The box suggests that the RX 7600 will also come with 8GB VRAM just like its predecessor.

Price is of course going to be the biggest factor for the RX 7600. AMD overpriced the 7900 XT by around $100 as it was launched at $899. However, the company has been forced to reduce the price due to poor sales and media reception. The GPU now often sells below $799, which is $100+ reduction in just six months.

However, sadly for AMD fans and consumers, Team Red might once again be making the same mistake. According to Cowcotland, the RX 7600 is allegedly priced at €349 at French retailers, which is roughly around US$329. If you remember, the RX 6600 too launched at $329 though back then, GPU cryptocurrency mining was massive, which led to shortages and higher prices.

For the purported MSRP of around ~$329, the RX 7600 is expected to offer performance similar to the often-overlooked RX 6700 10GB, though you are going to be getting 20% less VRAM, and that to for a similar price. So if you didn't care about power consumption, the 6700 10GB could be the better bet for you.

The RX 7600 is going to compete with Nvidia's RTX 4060, which will arrive after the 4060 Ti lands sometime at the end of this month according to media reports. The AMD RX 7600 will also be purportedly arriving around the same time as the RTX 4060 Ti, on May 25th. Hence, AMD hopes to have a head start at the $300-350 segment.

In terms of specifications, the RX 7600 will have 32 Compute Units (CUs) or 2048 Stream Processors, and the 8GB 20Gbps GDDR6 VRAM will be across a 128-bit wide interface. The memory bandwidth will add up to 320GB/s. It will have 32MB of Infinity (L3) Cache. The AMD RX 7600 is basically the desktop equivalent of the mobile RX 7600M XT, which was revealed at CES.

AMD prepping RX 7600 to take on Nvidia RTX 4060, and it too might be overpriced

ChatGPT's "helpful, honest, and harmless" AI rival Claude is now capable of reading, remembering, and analyzing 75,000 words in less than 60 seconds.

Anthropic announced(Opens in a new window) it has expanded Claude's context window from 9,000 to 100,000 tokens, which is roughly 75,000 words. It means entire books and long, dense documents can be fed into Claude very quickly and then the AI can be asked complex questions about them immediately, or provide instant summaries.

As a way of demonstrating how well that works, Anthropic asked Claude to read The Great Gatsby and then set about changing one line in the text to read, "a software engineer that works on machine learning tooling at Anthropic." The Great Gatsby is 72,000 tokens (54,000 words), and when asked to explain what had changed after the sentence tweak, Claude returned the correct answer in just 22 seconds.

This expanded context window isn't just for reading text, though. Claude can consume around six hours of audio using 100,000 tokens, which means it can easily transcribe podcasts and produce summaries from the contents of audio files without any work required by the user.

The use cases for an AI which can consume and understand such vast quantities of information so quickly are wide ranging—a programmer could ask Claude to read developer documentation to get answers to coding questions, dense legislation documents can be fed in and Claude can produce a list of the pros and cons, and research papers can be summarized into a more easily digestible form.

Meanwhile, Meta is using AI to generate text and images for use in ads, and Google is infusing its AI chatbot technology into its traditional search engine experience. Will all this focus on AI lead to a productivity boost, or just leave more of us looking for a new job?

Source

The issue is happening with Phison's PS5026-E26 controller which is fairly common in PCIe Gen5 drives. The recently reviewed Corsair MP700 was found to be shutting down as a result of overheating. The following charts from TechPowerUp show this behavior where the SSDs shut down due to the temperature shooting up.

Crucial's T700, which also uses the same controller is exhibiting similar issues, though it throttles heavily rather than completely shutting off. Others like the Gigabyte Gen5 10000 are likely to exhibit similar symptoms as it too uses the same E26 controller.

As you can see in the image below (via ComputerBase on Twitter), the T700 is severely underperforming in some of these tests.

Phison responded about the issue with the following statement to Phoronix:

After carefully reviewing the recent reports from TechPowerUp and Phoronix, Phison would like to acknowledge the issue found in the reviews of products using the new Phison PS5026-E26 controller. We take this matter seriously and are committed to resolving it promptly.

Our firmware engineering teams have already isolated the problem and made the necessary adjustments to the thermal throttle curve within hours of the report. However, the new firmware must undergo Phison's strict validation process before our partners can release it to customers. Rest assured our partners will notify end-users as soon as the validated update is available.

It is important to note that all E26 SSDs shipped without a heatsink are intended to be used with a heatsink. Most motherboards shipping with PCIe Gen5 enabled also include cooling specifically designed for Gen5 SSDs. We offer the 'bare drive' option to allow customers to use their existing cooling products.

We want to emphasize our commitment to providing high-quality products and solutions to our customers and will continue to work diligently to ensure their satisfaction. Thank you for your patience and understanding during this process.

Hence, Phison has made it clear that PCIe 5.0 SSDs are meant to be paired up with heatsinks only. The company had already warned about potential thermal issues back more than a year ago. In fact, Phison even suggested that it could require active cooling. Perhaps a slower PCIe 5.0 drive, that's based on lower speed memory, may not be a bad idea after all in case you don't have the room to add a heatsink.

Crucial, Corsair PCIe 5.0 SSDs running hot, throttling, and shutting down without heatsink

The announcement on the game's Steam page also came with a brief gameplay trailer showing off more of its graphics. If you are wondering if your gaming rig can handle it, Bloober Team has also posted its PC system requirements.

*(TSR= Temporal Super Resolution)

Finally you can check out the new cinematic intro for Layers of Fear. The full game is coming out sometime in June for the PC, PlayStation 5, and the Xbox Series X and S consoles.

Check out the PC hardware requirements for Layers of Fear before its Steam demo on Monday

This revamped version, set to roll out on May 11, promises to revolutionize how users interact with their smart devices. With a five-tab layout and improved functionality, controlling connected devices, automating home tasks, and gaining insights into home activities will be more seamless than ever.

Streamlined control with a five-tab layout

The highlight of the Google Home redesign is its visually appealing five-tab layout, which prioritizes user experience. This enhanced interface ensures effortless management of connected devices and simplifies home automation processes.

Among the notable tabs is the long-awaited "Favorites" section, allowing users to pin their most frequently used devices, actions, and automations for quick access. Additionally, the "Spaces" tab intelligently groups devices into categories such as lights, cameras, Wi-Fi, and climate control, enabling users to navigate their devices with ease.

Stay informed with the inbox

To ensure a seamless smart home experience, Google has introduced the "Inbox" feature within the redesigned Home app. This dedicated section acts as a central hub for crucial updates, notifying users about device issues and subscription reminders.

With the new and improved version of the Google Home app, users can expect an optimized user interface that enhances their overall smart home management.

Matter support for enhanced interoperability

In a commitment to fostering a more connected ecosystem, Google has expanded its support for Matter-enabled devices. This update empowers Google Home app users to seamlessly control over 60 different device types, ensuring effortless integration and compatibility.

Furthermore, this update also extends Matter support to iOS users, enabling iPhone users running iOS 16.5 to easily set up and manage Matter devices directly from the Google Home app.

Unlocking convenience for Pixel devices

Google is bringing added convenience to Pixel devices with the introduction of a new home panel. This innovative feature grants users quick access to the Google Home app directly from the lock screen or quick settings, eliminating the need to navigate through various menus.

Initially available for Pixel phones and later for Pixel Tablets, this home panel allows users to perform tasks such as adjusting thermostats, controlling lights, or monitoring cameras without unlocking their device or launching the app.

Expanded compatibility and accessibility

Google is expanding support for the Home app beyond just Android devices. The redesigned app will also be available for Wear OS-powered devices and Pixel Tablets, ensuring a seamless experience across platforms.

This expansion enables users to interact effortlessly with their connected devices, access important information, and enjoy an optimized smart home experience. Additionally, Nest camera users will benefit from improved accessibility, making it easier to view and manage camera feeds directly from within the Google Home app.

Google Home gets a brand-new interface

This week, at its annual I/O developer conference in Mountain View, Google showcased a head-spinning number of projects and products powered by or enhanced by AI. They included a new-and-improved version of its chatbot Bard, tools to help you write emails and documents or manipulate images, devices with AI baked in, and a chatbot-like experimental version of Google search. For a full recap of the event, complete with insightful and witty commentary from my WIRED colleagues, check out our Google I/O liveblog. 

Google’s big pivot is, of course, largely fueled not by algorithms but by generative AI FOMO.

The appearance last November of ChatGPT—the remarkably clever but still rather flawed chatbot from OpenAI—combined with Microsoft adding the technology to its search engine Bing a few months later, triggered something of a panic at Google. ChatGPT proved wildly popular with users, demonstrating new ways to serve up information that threatened Google’s vice grip on the search business and its reputation as the leader in AI.

The capabilities of ChatGPT and AI language algorithms like those powering it are so striking that some experts, including Geoffrey Hinton, a pioneering researcher who recently left Google, have felt compelled to warn that we might be building systems that we will someday struggle to control. OpenAI’s chatbot is often astonishingly good at generating coherent text on a given subject, summarizing information from the web, and even answering extremely tricky questions that require expert knowledge.

And yet, unfettered AI language models are also silver-tongued agents of chaos. They will gladly fabricate facts, express unpleasant biases, and say unpleasant or disturbing things with the right prompting. Microsoft was forced to limit the capabilities of Bing chat shortly after launch to avoid such embarrassing misbehavior, in part because its bot divulged its secret codename—Sydney—and accused a New York Times columnist of not loving his spouse.

Google worked hard to tone down the chaotic streak of text-generation technology as it prepared the experimental search feature announced yesterday that responds to search queries with chat-style answers synthesizing information from across the web.

Google’s smarter version of search is impressively narrow-minded, refusing to use the first person or talk about its thoughts or feelings. It completely avoids topics that might be considered risky, refusing to dispense medical advice or offer answers on potentially controversial topics such as US politics.

Google deserves recognition for reining in generative chatbots’ wild side like that. But in my tests, the new search interface felt incredibly tame compared to ChatGPT or Google’s own chatbot Bard.

As the company moves the technology into more of its products, perhaps the generative AI revolution will turn out to be a lot less fun than you might expect from the early shock and awe of ChatGPT, a chatbot that has an edgy charm. Gone are the wild ravings and imaginings of powerful AI bots. In their place are new ways to populate spreadsheets, compose email pleasantries, and find products to buy.

Even if “AI doomers” warning about errant AI prove overblown, it will be interesting to watch how companies like Google and OpenAI balance the development of more powerful generative language models with the need to have them behave.

Google has invested huge sums and major resources in AI over recent years, with CEO Sundar Pichai often pitching the company as “AI first”, and the company is desperate to show it can advance the technology more quickly than OpenAI. One high-level message from Google’s stream of AI announcements was that the company is not going to hold back anymore, as it did the LaMDA chatbot that was announced long before ChatGPT appeared but not made public.

In March, some big names in AI research signed an open letter calling for a six-month pause on creating machine learning systems more powerful than GPT-4, which powers ChatGPT. Pichai was not a signatory and said in his keynote speech yesterday that the company is currently training a new, more powerful language model called Gemini.

A source at Google tells me this new system will incorporate a range of recent advances from different large language models and may eclipse GPT-4. But don’t expect to get to experience the full power or charisma Gemini can offer. If Google applies the same chaos-taming methods seen in its chat-like search experiment, it may just seem like another surprisingly clever autocomplete.

The Boring Future of Generative AI

(May require free registration to view)

In a post on his Twitter account, Elon Musk says he has "hired a new CEO for X/Twitter. She will be starting in ~6 weeks!" That unnamed person likely will also be taking over as CEO of Twitter's parent company X. Musk will still be around at Twitter. He says his new role will be as "exec chair & CTO, overseeing product, software & sysops."

During his tenure as CEO of Twitter, Musk started by firing over half of its employees. He also made a lot of changes designed to charge people and businesses money for services, such as the $8 a month Twitter Blue subscription with a number of extra features, and charging most businesses money to access its APIs. This week Twitter started offering encrypted DMs for Twitter Blue and paid business users.

Elon Musk says he's hired a new Twitter CEO but he will still be around

They hope it can assist people who are paralyzed or suffer disability

Mind-reading technology can now transcribe people’s thoughts in real-time based on the blood flow in their brain.

A study put three people in MRI machines and got them to listen to stories.

For the first time, researchers claim, they produced a rolling text of people’s thoughts, and not just single words or sentences, without using a brain implant.

The mind-reading technology did not exactly replicate the stories, but captured the main points.

The breakthrough raises concerns about ‘mental privacy’ as it could be the first step in being able to eavesdrop on others’ thoughts.

Using technology similar to ChatGPT, the technology also interpreted what people were seeing when they watched silent films, or their thoughts as they imagined telling a story.

Researchers exposed patients to thousands of hours of podcasts to determine a link between words

and brain activity. Then, they asked the participants to either tell a story or recap and react to a news article

Researchers found that they could be able to read a person's thoughts with around 50 percent accuracy

by using a novel MRI scanning method. They would expose a patient to hours of podcasts to see how

their brain would react to different words, and then use the scan to determine 

The brain reacts to different words using electrical signals and blood flow. Researchers

built an AI model that can read those reactions and translate them into writing (file photo)

But the researchers point out that it took 16 hours of training, with someone listening to podcasts in an MRI machine, for the computer model to understand their brain patterns and interpret what they were thinking.

People were also able to ‘sabotage’ the technology, using methods like mentally listing animals’ names, to stop it reading their thoughts.

Jerry Tang, lead author of the study from the University of Texas at Austin, said he could not prove ‘a false sense of security’ that the technology might not have the potential to eavesdrop on people’s thoughts in the future, and said it could be ‘misused’ now.

But he said: ‘We take very seriously the concerns that it could be used for bad purposes.

‘And we want to dedicate a lot of time moving forward to try to try to avoid that.

‘I think, right now, while the technology is in such an early state, it's important to be proactive and get a head-start on, for one, enacting policies that protect people's mental privacy, giving people a right to their thoughts and their brain data.

‘We want to make sure that people only use these when they want to, and that it helps them.’

Silicon Valley is very interested in mind-reading technology which could one day allow people to type just by thinking the words they want to communicate.

Elon Musk’s company, Neuralink, is working on a brain implant which could provide direct communication with computers.

But the new technology is relatively unusual in its field, in reading thoughts without using any kind of brain implant, so that no surgery is required.

While currently, it requires a bulky, expensive MRI machine, in the future people might wear patches on their head which use waves of light to penetrate into the brain and provide information on blood flow.

This could allow people’s thoughts to be detected as they move around.

The new study, reported in the journal Nature Neuroscience, used a ‘decoder’, which incorporates both a computer model to interpret people’s brain activity and language-processing technology similar to ChatGPT to help generate possible words.

After hours of training, the decoder was able to pick up what people were thinking about half the time.

That meant it produced text which closely, and sometimes exactly, matched the words people were listening to - working this out using only their brain activity.

For example, one person listening to a speaker say ‘I don’t have my driver’s license yet’ had their thoughts translated as ‘she has not even started to learn to drive yet’.

The researchers say the breakthrough could help people who are mentally aware yet unable to speak, like stroke victims or those with motor neurone disease.

Unlike other mind-reading technology, it works when people think of any word, and not just those from a set list — although it struggles with pronouns like ‘he’ and ‘I’.

It detects activity in language-forming regions of the brain, unlike other similar technology which typically detects how someone is imagining moving their mouth to form specific words.

Dr Alexander Huth, senior author of the study from the University of Texas at Austin, said: ‘We were kind of shocked that this worked as well as it does.

‘This is a problem that I’ve been working on for 15 years.’

He added: ‘For a non-invasive method, this is a real leap forward compared to what’s been done before, which is typically single words or short sentences.’

On concerns that the technology could be used on someone without them knowing, such as by an authoritarian regime interrogating political prisoners or an employer spying on employees, the researchers say the system can only read the thoughts of an individual after being trained on their thought patterns, so could not be applied to someone secretly.

Dr Huth said: ‘If people don't want to have something decoded from their brains, they can control that using just their cognition - they can think about other things, and then it all breaks down.’

SOURCE:

https://www.dailymail.co.uk/sciencetech/article-12033319/Scientists-read-MIND-AI-turns-peoples-thoughts-text-real-time.html

in responding to patient questions

An artificial intelligence chatbot was able to outperform human doctors in responding to patient questions posted online, according to evaluators in a new study.

Research published in the Journal of the American Medical Association (JAMA) Internal Medicine found that a chatbot’s responses to patient questions, pulled from a social media platform, were rated “significantly higher for both quality and empathy.”

Researchers from a number of institutions, including the University of California San Diego, Bryn Mawr College and Johns Hopkins University, presented a team of licensed health care professionals with responses to 195 randomly drawn patient questions. The evaluators determined that patients preferred the chatbot responses.

Of the 195 questions and responses — which were reviewed in triplicate for a total of 585 evaluations — evaluators preferred chatbot responses to physician responses in 78.6 percent of the cases.

But despite the study’s “promising results” on the artificial intelligence tech for patient questions, the researchers stressed that it’s “crucial to note that further research is necessary before any definitive conclusions can be made regarding their potential effect in clinical settings.”

The study suggests that, after further study, chatbots could be used to draft responses to patient questions that physicians could then edit.

More than 30 percent of LGBTQ youth attribute poor mental health to anti-LGBTQ laws

“The rapid expansion of virtual health care has caused a surge in patient messages concomitant with more work and burnout among health care professionals. Artificial intelligence (AI) assistants could potentially aid in creating answers to patient questions by drafting responses that could be reviewed by clinicians,” the researchers said.

Researchers have been looking into the rapidly emerging tech and grappling with how it could impact different sectors, as controversy swirls over use of the tech in settings like school.

Earlier this year, a study found ChatGPT could pass an exam at the Wharton Business School. The chatbot was later able to score in the top 10 percent of test-takers in a simulated bar exam.

SOURCE

https://thehill.com/policy/technology/3982046-ai-chatbot-outperforms-human-doctors-in-responding-to-patient-questions/
 

A European Parliament committee voted to strengthen the flagship legislative proposal as it heads toward passage, part of a yearslong effort by Brussels to draw up guardrails for artificial intelligence. Those efforts have taken on more urgency as the rapid advances of chatbots like ChatGPT highlight benefits the emerging technology can bring — and the new perils it poses.

Here’s a look at the EU’s Artificial Intelligence Act:

HOW DO THE RULES WORK?

The AI Act, first proposed in 2021, will govern any product or service that uses an artificial intelligence system. The act will classify AI systems according to four levels of risk, from minimal to unacceptable. Riskier applications will face tougher requirements, including being more transparent and using accurate data. Think about it as a “risk management system for AI,” said Johann Laux, an expert at the Oxford Internet Institute.

WHAT ARE THE RISKS?

One of the EU’s main goals is to guard against any AI threats to health and safety and protect fundamental rights and values.

That means some AI uses are an absolute no-no, such as “social scoring” systems that judge people based on their behavior. AI that exploits vulnerable people including children or that uses subliminal manipulation that can result in harm, such as an interactive talking toy that encourages dangerous behavior, is also forbidden.

Lawmakers beefed up the proposal by voting to ban predictive policing tools, which crunch data to forecast where crimes will happen and who will commit them. They also approved a widened ban on remote facial recognition, save for a few law enforcement exceptions like preventing a specific terrorist threat. The technology scans passers-by and uses AI to match their faces to a database.

The aim is “to avoid a controlled society based on AI,” Brando Benifei, the Italian lawmaker helping lead the European Parliament’s AI efforts, told reporters Wednesday. “We think that these technologies could be used instead of the good also for the bad, and we consider the risks to be too high.”

AI systems used in high risk categories like employment and education, which would affect the course of a person’s life, face tough requirements such as being transparent with users and putting in place risk assessment and mitigation measures.

The EU’s executive arm says most AI systems, such as video games or spam filters, fall into the low- or no-risk category.

WHAT ABOUT CHATGPT?

The original 108-page proposal barely mentioned chatbots, merely requiring them to be labeled so users know they’re interacting with a machine. Negotiators later added provisions to cover general purpose AI like ChatGPT, subjecting them to some of the same requirements as high-risk systems.

One key addition is a requirement to thoroughly document any copyright material used to teach AI systems how to generate text, images, video or music that resembles human work. That would let content creators know if their blog posts, digital books, scientific articles or pop songs have been used to train algorithms that power systems like ChatGPT. Then they could decide whether their work has been copied and seek redress.

WHY ARE THE EU RULES SO IMPORTANT?

The European Union isn’t a big player in cutting-edge AI development. That role is taken by the U.S. and China. But Brussels often plays a trendsetting role with regulations that tend to become de facto global standards.

“Europeans are, globally speaking, fairly wealthy and there’s a lot of them,” so companies and organizations often decide that the sheer size of the bloc’s single market with 450 million consumers makes it easier to comply than develop different products for different regions, Laux said.

But it’s not just a matter of cracking down. By laying down common rules for AI, Brussels is also trying to develop the market by instilling confidence among users, Laux said.

“The thinking behind it is if you can induce people to to place trust in AI and in applications, they will also use it more,” Laux said. “And when they use it more, they will unlock the economic and social potential of AI.”

WHAT IF YOU BREAK THE RULES?

Violations will draw fines of up to 30 million euros ($33 million) or 6% of a company’s annual global revenue, which in the case of tech companies like Google and Microsoft could amount to billions.

WHAT’S NEXT?

It could be years before the rules fully take effect. European Union lawmakers are now due to vote on the draft legislation at a plenary session in mid-June. Then it moves into three-way negotiations involving the bloc’s 27 member states, the Parliament and the executive Commission, where it could face more changes as they wrangle over the details. Final approval is expected by the end of the year, or early 2024 at the latest, followed by a grace period for companies and organizations to adapt, often around two years.

Source

• Shares of Alphabet are trading at their highest level since August on the heels of Google’s artificial intelligence announcements at its developer conference Wednesday.

• At its I/O event, Alphabet-owned Google announced it would bring AI features to its signature search product to turn complex queries into simple answers by synthesizing results.

• Analysts seemed pleased that Google’s announcements will keep it at the forefront of the AI race and generate returns for the company.

Shares of Alphabet are trading at their highest level since August on the heels of Google’s artificial intelligence announcements at its developer conference Wednesday.

The stock was up about 5% and trading near $118 per share as of Thursday morning. It last traded above that price in August last year.

At its I/O event, Alphabet-owned Google announced it would bring AI features to its signature search product to turn complex queries into simple answers by combining results from multiple sources.

Google also said its suite of workplace tools like Google Docs would soon let users create new documents and fill out spreadsheets with AI. The company also said it would make its Bard AI chatbot more widely available and that it would soon have the ability to respond in different languages and images.

Analysts seemed pleased that Google’s announcements will keep it at the forefront of the AI race and generate returns for the company.

“We remain bullish and view I/O as a clearing event that will help erode the heavily debated ‘AI overhang’…and narrow GOOGL’s current valuation gap vs. peers,” Morgan Stanley analyst Brian Nowak wrote.

Source

It published an interim order stating they would need "prior written consent" from the UK regulator before making acquisitions between the companies.

The order applies to the businesses as well as their subsidiaries.

It comes two weeks after the regulator blocked Microsoft's $68.7bn (£54.8bn) deal to buy Activision Blizzard.

The takeover would have seen Microsoft acquire such hit titles as Call of Duty, Overwatch and World of Warcraft.

But the regulator said it was concerned the deal would offer reduced innovation and less choice for gamers in the cloud gaming business.

Microsoft and Activision criticised the decision and said they would appeal, with an Activision official saying at the time it showed the UK was "clearly closed for business".

To go through, the deal has to be approved by regulatory bodies in the UK, United States and European Union.

The CMA was the first regulator to make a ruling, in April. The EU is expected to make a decision in May.

In the interim order, the CMA wrote it was "preventing pre-emptive action" from Microsoft or Activision Blizzard.

The order prevents the businesses from acquiring an interest in each other, including their subsidiaries, or businesses that themselves have interest in the companies.

For example:

• Activision Blizzard could not invest in Microsoft's Xbox Game Studios

• Microsoft could not invest in Activision Blizzard subsidiaries such as King, the studio that makes hit mobile game Candy Crush Saga

The order states the companies should "immediately notify the CMA" if they have "any reason to suspect" the order has been breached.

A Microsoft official told BBC News: "We remain firmly committed to this deal and look forward to presenting our case to the Competition Appeal Tribunal."

Activision Blizzard has been approached for comment.

Source

The European Space Agency’s (ESA) Euclid satellite completed the first part of its long journey into space on May 1 2023, when it arrived in Florida on a boat from Italy. It is scheduled to lift off on a Falcon 9 rocket, built by SpaceX, from Cape Canaveral in early July.

Euclid is designed to provide us with a better understanding of the “mysterious” components of our universe, known as dark matter and dark energy.

Unlike the normal matter we experience here on Earth, dark matter neither reflects nor emits light. It binds galaxies together and is thought to make up about 80% of all the mass in the universe. We’ve known about it for a century, but its true nature remains an enigma.

Dark energy is similarly puzzling. Astronomers have shown that the expansion of the universe over the last five billion years has been accelerating faster than expected. Many believe this acceleration is driven by an unseen force, which has been dubbed dark energy. This makes up about 70% of the energy in the universe.

Euclid will map this “dark universe”, using a suite of scientific instruments to shed light on different aspects of dark energy and dark matter.

A light in the dark

After launch, Euclid will undertake a month-long journey to a region in space called the second Earth-Sun Lagrangian point, which is five times further from us than the Moon. It’s where the gravitational pull of the Sun and the Earth balance out and provides a stable vantage point for Euclid to observe the universe. Euclid will join the James Webb Space Telescope (JWST) at this point and will be the perfect companion to that amazing space observatory.

My involvement in Euclid began in 2007 when I was invited by ESA to participate in an independent concept advisory team to assess two competing mission proposals called SPACE and DUNE.

Both used different techniques, and therefore different instruments, to study the dark universe, and ESA was struggling to decide between them. Both were compelling concepts and our team decided that both had merit, especially to provide a vital cross-check between them. Euclid was thus born from the best of both concepts.

Euclid is designed to study the whole universe so needs instruments with wide fields of view. The wider the field of view of the imaging instrument, the more of the universe it can observe. To do this, Euclid uses a relatively small telescope compared to JWST. In size, Euclid is roughly the size of a truck compared to the aircraft-sized JWST. But Euclid also carries some of the biggest digital cameras deployed in space with fields of view hundreds of times greater than JWST’s.

Shapes and colours

The Euclid VIS (or visible) instrument, built mostly in the UK, is designed to measure the positions and shapes of as many galaxies as possible to look for subtle correlations in this data caused by the gravitational lensing of the light, as it travels to us through the intervening dark matter. This gravitational lensing effect is weak, only one part in a hundred thousand for most galaxies, thus requiring lots of galaxies to see the effect in high definition. Thus VIS will produce Hubble telescope-like image quality over a third of the night sky.

VIS, however, can’t measure the colours of objects. This is needed to measure their distance through the redshift effect, where light from those objects is shifted to longer, or redder, wavelengths in a way that relates to their distance from us. Some of this data will need to come from existing and planned ground-based observatories, but Euclid also carries the NISP (Near-Infra Spectrometer and Photometer) instrument which is specifically designed to measure the infrared colours and spectra, and therefore redshifts, for the most distant galaxies that Euclid will see.

To measure dark energy, NISP will exploit a relative new technique called Baryon Acoustic Oscillations (BAO) that provides an accurate measurement of the expansion history of the universe over its last 10 billion years. That history is vital for testing possible models of dark energy including suggested modifications to Einstien’s Theory of General Relativity.

Euclid will gather information on the shapes and other properties of galaxies in the sky. Image Credit: NASA, ESA, S. Beckwith (STScI), and The Hubble Heritage Team (STScI/AURA), CC BY 4.0

Treasure trove

Such an experiment takes an army of scientists and not everyone is solely working on dark matter and dark energy. Like JWST, Euclid will be a treasure-trove of new discoveries in many areas of astronomy. The Euclid consortium needs hundreds of people to help develop the sophisticated software needed to merge the space data with the ground-based data, and extract, to high accuracy, the shapes and colours of billions of galaxies.

This software has also been checked and verified using some of the largest simulations of the universe that have ever been constructed. After arriving at L2, Euclid will undergo several months of testing, validation and calibration to ensure the instruments and telescope are working as expected. We are all familiar with such nervous waiting after the recent JWST launch.

Once ready, Euclid will embark on a five-year survey of 15,000 square degrees of the sky with about 2,000 scientists from across the world collecting results along the way. However, the true power of Euclid will only be realised once we have all this data together and analysed carefully. That could take another five years, taking us well into next decade before we have our final dark answers. The SpaceX launch therefore only feels like the half-way point in the Euclid story.

I will travel to Florida this summer to see the launch of Euclid. I will be joined by hundreds of my colleagues who have dedicated their careers to building this amazing telescope and experiment. Seeing the project come together in this way makes me proud to call myself a “Euclidian”.

Robert Nichol, Pro Vice-Chancellor and Executive Dean, University of Surrey

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Source

In the days leading up to its annual software conference, Google executives were unusually excited about … wallpaper. Phone wallpaper.

Wallpaper that changes and breathes and undulates on Android phones, specifically. Wallpaper that’s generated by artificial intelligence after tapping a few prompts on your phone screen.

“Generative AI” is the key phrase here. It’s the category of artificial intelligence that tech companies both big and small are hanging their futures on right now. Alphabet-owned Google is one of the pioneers in this space; as Google executives like to remind people, the “T” in OpenAI’s ChatGPT actually refers to transformer technology that Google introduced back in 2017. And Google has been working on humanlike chatbot technology for years.

But now Google is largely perceived as being behind in generative AI, since both OpenAI and Microsoft (which funds OpenAI) released their GenAI chat tools before Google had a chance to unleash its own tool, called Bard. The release of ChatGPT even reportedly triggered a “code red” for Google, leading the company to funnel more resources into building AI products, and yesterday.

So it makes sense that Google would use I/O, its annual developer conference that kicked off today in Mountain View, California, as a stage to tout its latest AI features. AI is the thread woven throughout the program. It’s also featuring prominently in two new Android-specific features, demonstrating just how eager Google is to unleash usable products around generative AI and get into the hands of potentially billions of people.

“We’re in this position where Google has pioneered much of the seminal research, particularly around this class of large language models, which is the basis of this nonlinear step forward in tech,” Dave Burke, vice president of engineering for Android, said in an interview with WIRED. “And at the same time we have these large surfaces of Android. And we can bring the two things together.”

“For Google, the Android ecosystem offers a platform to prove its innovative AI portfolio at a scale that not all of their cloud peers can compete with,” says Chirag Dekate, vice president and analyst at Gartner.

Picture This

The new suite of generative wallpapers for Android seem basic but use varying levels of AI. There’s a customizable emoji option and a “cinematic” mode that turns still photos into mini movies. The latter uses on-device machine learning to determine what’s foreground in the image and what’s background, then creates a subtly zooming parallax effect. Generative AI wallpaper is more, well, generative: The user will tap on a series of predetermined prompts offered on the phone, like “classic arts,” “City by the bay,” and “Post-Impressionist,” which then generates entirely new and original wallpapers.

Google says its image-generation model is trained on public domain artwork, which means it’s likely hoping to head off any sticky ethical issues around the AI art copyright. It also won’t be available to the masses right away: The wallpapers will roll out only to Google’s own Pixel phones first, starting next month.

So … wallpaper? Google’s seminal research has resulted in AI Van Gogh on your phone? Well, yes, but Burke insists that while “it might seem obvious or simple in retrospect, this is one of the ways we’re thinking that AI can be applied in a very user-centric, responsible way.”

“We’ve been doing things with AI for a long time as a company. And the advancements in generative AI are pretty exciting,” says Sameer Samat, vice president of product management for Android and the Google Play app store. “But these aren’t that old in terms of putting them into products. So when you talk about productizing all of this, it’s extremely early days.”

Computer Talk

Another area of Android getting the GenAI treatment: messages. This summer, Google’s Messages app will be imbued with a Magic Compose feature, which uses generative AI to offer suggested responses based on the context of your messages. A lot of messaging apps already do a version of this by predicting which word you might want to use next, but Magic Compose will generate entire sentences (and will helpfully suggest them before you hit Send).

Magic Compose can also change the tone of your message, if you prompt it to do so, making something sound more formal, more poetic, or more like Shakespeare (really). Burke says this is built on what’s known as a few-shot training model, in which a model can learn to perform an app-specific function based on just a handful of training examples.

Ben Bajarin, chief executive and principal analyst at Creative Strategies, says Google’s limited rollout of generative AI on Android is a smart approach. “At companies like Apple, and even at Microsoft, there is a genuine concern around this technology. So Google seems to be erring on caution and making sure this is trustworthy before bringing it to billions of customers,” he says.

Bajarin also pointed out the computing limitations that still exist around these technologies. “On-device AI requires a tremendous amount of computing power,” he says. “Rolling it out to native software will require using the local processing power and not the cloud. If you look at some of the basic demos of Stable Diffusion that Qualcomm has shown, it really highlights the lack of native computing power on our mobile devices.”

Android’s new generative AI and customization features are just part of a much broader software update. Android 14, which started rolling out as beta software in February, will also include support for larger screens (like the Pixel Fold), battery optimization features, and enhanced security features, like password-replacing passkeys. Google said today that it counts more than 3 billion active Android devices around the globe, spanning phones, tablets, cars, and TVs. On the TV front, Google claims that Android TV OS is the top streaming platform in the world by shipments. And WearOS, its wearable devices software, is fast-growing.

On stage today, Samat emphasized interconnectivity across different hardware devices, whether that’s by giving people the ability to cast media to non-Google hardware or to find their Apple AirPods more easily from their Pixel phone. Samat noted that more than 800 million phones now use RCS messaging, a standard that Google supports—and that Apple has famously withheld support for. Google executives also reminded the I/O crowd today that the company is working with Samsung to build an Android-based XR system. If you can’t beat ’em, join them?

Unless it’s generative AI products. Then, if you’re Google, the pressure is on.

Source

Google is expanding and modernizing its Find My Device platform, bringing it closer in line with location-tracking solutions from Apple and Tile. During today’s I/O 2023 keynote, Google’s Sameer Samat announced that headphones / earbuds, tablets, and other product categories will be added to Find My Device in the coming months. (Some of these are already supported, but you can count on seeing many additional devices.) More importantly, it will now become a vast network that uses devices owned by other people to ping your missing gadget and pinpoint its location.

“It’s powered by a network of billions of Android devices around the world,” Samat said. Tile, Chipolo, and other third parties will participate in the updated program. “From the start, we designed the network in a privacy-preserving way,” the Google VP added, noting that device location data is encrypted and not viewable by the company.

Samat also addressed the potential for malicious use of these location-tracking products. Unknown tracker alerts will reveal when a tracking device is traveling with you — including Apple AirTags — and he mentioned how the two companies are working together to prevent stalking with these gadgets intended for convenience.

Source

According to Seoul Metro, the operator of the subway service in Seoul, the new service will help analyze the real-time flow of passengers and crowd levels in subway compartments, improving operational efficiency.

Currently, the subway operator monitors crowding and measures the number of passengers per compartment at Line Nos. 1-8 by tracking traffic card information at turnstiles. It also weighs each subway section using load cells and real-time tracking of passengers’ mobile phone data usage to determine the number of passengers on board. Based on the collected information, the train operator then tries to reduce passenger flows in subways by dispersing them to other compartments, the operator said.

In response to rising concerns from subway riders about receiving incorrect data on congestion, especially during rush hours, Seoul Metro said that the service will enhance real-time data accuracy in crowded places.

According to data compiled by the subway operator, the average congestion levels at Line Nos. 1-8 stood at 145.7 percent last year, up 11.5 percent from 134.2 percent in 2019. Line No. 4 had the highest congestion level of 185.8 percent in 2022, followed by Line No. 2 at 172.3 percent, and Line No. 7 at 160.6 percent.

With the upgrade, the subway operator said it hopes to better monitor the level of crowded spaces at subway platforms, stairways and indoor waiting rooms.

“The system will create reliable data, help handle congestion levels and improve such situations,” an official at Seoul Metro said via a press release.

The subway operator plans to inject 500 million won ($377,000) into developing and completing the traffic flow management system by this year’s end to provide efficient services for passengers. The subway operator will then evaluate the reliability of the system through a series of tests during a system evaluation process implementation.

As part of its initiative to streamline subway operations, the subway operator has installed a moveable subway fence at City Hall Station on Line Nos. 1 and 2, to allow passengers to move in one direction and to reduce the chance of collisions. A safety fence that separates sections on staircases at Seoul National University Station on Line No. 2 has been also installed, according to the metro operator.

Source

Twitter has been in the news a lot these days since Elon Musk took over as the owner of the social media platform by purchasing it for $43 billion about a year ago.

From some advertising companies leaving and possibly some of them rejoining Twitter to advertise their products. To talks of Twitter unbanning some previously banned accounts in the name of freedom of speech. A lot of changes are happening on the social media platform.

Recently, Twitter also made some API changes, making it harder for sites like ours to freely auto-post our articles on Twitter.

Not to forget, Twitter’s famous episode where one now needs to pay Twitter $8 per month to get a blue check mark, which the platform calls “Twitter Blue”. This feature also allows one to post longer tweets and edit already posted ones, among many other things.

Now it seems that Twitter is introducing even more features to make it more attractive to the users.

Elon Musk has made an interesting tweet announcing new features for Twitter:

In it, he mentions a few things.

Encrypted Direct Messages On Twitter

In the tweet, Elon Musk mentions that in the latest version of the app, the users can direct message a reply to any message in a Twitter thread.

Secondly, he says that starting tomorrow, what he calls encrypted DMs (direct messages) V1.0, would be released to the public.

Interestingly, he reveals that ideally, even he himself shouldn’t be able to read direct messages no matter what.

While Twitter calls this move encrypted DMs V1.0, we wonder what would happen when Twitter launches encrypted DMs V2.0.

Voice And Video Chat On Twitter

Another important thing Elon Musk has announced is that Twitter will soon add a support for voice and video calls on Twitter.

Elon Musk said in the tweet that one can call anyone from their own Twitter handle. Through it, he claims one can call anyone in the world without need to give the other side your phone number. Direct Twitter To Twitter calls.

Interestingly, there’s no mention if voice and video calls would be encrypted too.

Welcome Moves

Both the moves announced by Twitter should be welcomed.

Encrypting messages gives privacy to users. Not to forget voice and video calling. This feature is not new, other social media platforms have introduced it before too. However, Twitter introducing it means it’s directly challenging the likes of Zoom, Meet and even WhatsApp.

For now, we don’t know if voice and video calls would be available for desktop (unlikely) or will they be limited to just mobile apps.

Nor do we know how Twitter intends to counter the spam that might come with it. Whether it would be allowed only for paid users or for everyone. If it’s enabled for everyone, then the likelihood of spam calls might increase further. If so, then Twitter might need to take a lot of measures to stop them.

We might hear more about it when it gets rolled out for its users.

Twitter To Introduce Voice And Video Calling On The Platform

Google's blog post states that it will test these new features in Search Labs, before they roll out to all users. Using generative AI will allow users to type in more complete search prompts. It stated:

Let’s take a question like “what's better for a family with kids under 3 and a dog, bryce canyon or arches.” Normally, you might break this one question down into smaller ones, sort through the vast information available, and start to piece things together yourself. With generative AI, Search can do some of that heavy lifting for you. You’ll see an AI-powered snapshot of key information to consider, with links to dig deeper.

The new AI features will also include more details when shopping for an item. Google states:

When searching for a product, you’ll get a snapshot of noteworthy factors to consider and products that fit the bill. You’ll also get product descriptions that include relevant, up-to-date reviews, ratings, prices and product images.

In another blog post, Google also talked about a new Perspectives feature for Search that is designed to help people learn about something from the thoughts of others. It stated:

Let’s say you’re moving across the country, and you don’t know anyone who lives there yet. You search for “how to make friends in a new city,” and tap the Perspectives filter, which shows you a page of results with advice from other people, like personal stories told through video, or tips from commenters in a forum thread.

Stay tuned as we report more news from Google I/O today.

Google is "supercharging" its classic Search platform with AI for info, shopping and more

For this year's limited-time experiences, Google is offering the following experiments in their Labs:

• SGE ( Search Generative Experience): As we have seen with the Bing chatbot, with new capabilities brought by generative AI, Google plans to make Search easier for the masses. Its Bard AI will understand the gist of the search query and provide users with a more informative and useful response.

• Code Tips: It will give short useful code extracts using the power of LLM. Currently supported languages are Go, Java, Python, Javascript, C++, Kotlin, shell, Git and Docker.

• Add to Sheets: While hovering over a search result, Google will show add to sheets icon, bookmark, and share options. This will directly add the link to a selected spreadsheet on Google Drive.

These are the three currently offered experiments on Search Labs. It's open today in English, US-only. Sign-up for the waitlist is required before getting access. You can sign up by following this link.

Source: Google

Google asks users to help build the future of Search with Search Labs

Google stated in a blog post that Duet AI will help Workspace users in many different apps. Here's an example of how it will work in Google Slides:

A picture is worth a thousand words, but until now, creating unique and compelling visuals for presentations has been a manual and time-consuming process. We’re embedding Duet AI into Slides so you can easily generate images with a few words — the real power of these image models is that they can visualize something that has never existed. Perhaps you’re a marketer doing early concept brainstorming with your creative agency on a light-hearted campaign to get Parisians to go on safari and want to provide clear visual input early in the process to prevent wasted work later on. Now, you can generate an original visual that conveys your unique artistic vision, all from a simple prompt.

Duet AI is also coming for Gmail on mobile to help compose emails, check out data in Slides, and create better online meetings in Google Meet. You can sign up for the waitlist to check out Duet AI.

Google Workspace to add Duet AI for virtual collaboration for Gmail, Slides and more