A machine-learning tool can easily spot when chemistry papers are written using the chatbot ChatGPT, according to a study published on 6 November in Cell Reports Physical Science. The specialized classifier, which outperformed two existing artificial intelligence (AI) detectors, could help academic publishers to identify papers created by AI text generators.
“Most of the field of text analysis wants a really general detector that will work on anything,” says co-author Heather Desaire, a chemist at the University of Kansas in Lawrence. But by making a tool that focuses on a particular type of paper, “we were really going after accuracy.”
The findings suggest that efforts to develop AI detectors could be boosted by tailoring software to specific types of writing, Desaire says. “If you can build something quickly and easily, then it’s not that hard to build something for different domains.”
The elements of style
Desaire and her colleagues first described their ChatGPT detector in June, when they applied it to Perspective articles from the journal Science. Using machine learning, the detector examines 20 features of writing style, including variation in sentence lengths, and the frequency of certain words and punctuation marks, to determine whether an academic scientist or ChatGPT wrote a piece of text. The findings show that “you could use a small set of features to get a high level of accuracy,” Desaire says.
In the latest study, the detector was trained on the introductory sections of papers from ten chemistry journals published by the American Chemical Society (ACS). The team chose the introduction because this section of a paper is fairly easy for ChatGPT to write if it has access to background literature, Desaire says. The researchers trained their tool on 100 published introductions to serve as human-written text, and then asked ChatGPT-3.5 to write 200 introductions in ACS journal style. For 100 of these, the tool was provided with the papers’ titles, and for the other 100, it was given their abstracts.
When tested on introductions written by people and those generated by AI from the same journals, the tool identified ChatGPT-3.5-written sections based on titles with 100% accuracy. For the ChatGPT-generated introductions based on abstracts, the accuracy was slightly lower, at 98%. The tool worked just as well with text written by ChatGPT-4, the latest version of the chatbot. By contrast, the AI detector ZeroGPT identified AI-written introductions with an accuracy of only about 35–65%, depending on the version of ChatGPT used and whether the introduction had been generated from the title or the abstract of the paper. A text-classifier tool produced by OpenAI, the maker of ChatGPT, also performed poorly — it was able to spot AI-written introductions with an accuracy of around 10–55%.
The new ChatGPT catcher even performed well with introductions from journals it wasn’t trained on, and it caught AI text that was created from a variety of prompts, including one aimed to confuse AI detectors. However, the system is highly specialized for scientific journal articles. When presented with real articles from university newspapers, it failed to recognize them as being written by humans.
Wider issues
What the authors are doing is “something fascinating,” says Debora Weber-Wulff, a computer scientist who studies academic plagiarism at the HTW Berlin University of Applied Sciences. Many existing tools try to determine authorship by searching for the predictive text patterns of AI-generated writing rather than by looking at features of writing style, she says. “I’d never thought of using stylometrics on ChatGPT.”
But Weber-Wulff points out that there are other issues driving the use of ChatGPT in academia. Many researchers are under pressure to quickly churn out papers, she notes, or they might not see the process of writing a paper as an important part of science. AI-detection tools will not address these issues, and should not be seen as “a magic software solution to a social problem.”
This article is reproduced with permission and was first published on January 27 2023.
A Parkinson’s patient can now walk 6km (3.7 miles) thanks to an implant targeting the spinal cord. The Guardianreports that the man — 62-year-old “Marc” from Bordeaux, France — developed severe mobility impairments from the degenerative disease. “I practically could not walk anymore without falling frequently, several times a day,” he said in a press release announcing the breakthrough. “In some situations, such as entering a lift, I’d trample on the spot, as though I was frozen there, you might say.” Wearing the spinal implant allows him to walk “almost normally” as the research team eyes a full clinical trial.
Marc underwent a “precision neurosurgical procedure” two years ago at Lausanne University Hospital (CHUV), which helped facilitate the research. The surgery fitted him with an electrode field placed against his spinal cord and an electrical impulse generator under the skin of his abdomen. Although conventional Parkinson’s treatments often target brain regions affected by the loss of dopamine-producing neurons, this approach instead focuses on the spinal area associated with activating leg muscles for walking.
The procedure used a personalized map of Marc’s spinal cord, identifying the specific locations signaling leg movements. He wears a movement sensor on each leg that tells the implant he’s trying to walk; it then switches on and sends electrical impulses to the targeted spinal neurons, adapting to his movement in real-time.
GABRIEL MONNET via Getty Images
“In response to precise stimulation of the lumbar spinal cord, I’ve observed for the first time remarkable improvements of gait deficits due to Parkinson’s disease,” project supervisor Jocelyne Bloch, professor and neurosurgeon at CHUV Lausanne University hospital, said in a webinar discussing the patient’s success. “I really believe that these results open realistic perspectives to develop a treatment.”
The patient says he could walk practically normally with the stimulation after several weeks of rehab. He now wears it for around eight hours daily, only turning it off when sleeping or lying down for a while. “I turn on the stimulation in the morning and I turn off in the evening,” he said. “This allows me to walk better and to stabilise. Right now, I’m not even afraid of the stairs anymore. Every Sunday I go to the lake, and I walk around 6 kilometres. It’s incredible.”
The researchers caution that there’s still a vast chasm between tailoring the approach to one person vs. optimizing it for wide-scale use. Co-leads Grégoire Courtine and Bloch are working on a commercial version of the neuroprosthetic in conjunction with Onward Medical. “Our ambition is to provide general access to this innovative technology to improve the quality of life of Parkinson’s patients significantly, all over the world,” they said.
Michael J. Fox (right) with Sting.
Michael J. Fox Foundation
In the meantime, research on six new patients will continue in 2024. The team says a “generous donation” of $1 million from the Michael J. Fox Foundation for Parkinson’s Research is funding the upcoming work. In 2021, the actor’s organization announced it had contributed over $1.5 billion to Parkinson’s research.
This article originally appeared on Engadget at https://ift.tt/GkTKEeQ
Tesla unveiled its Optimus humanoid robot prototype in 2021. It remains to be seen whether China’s plans could rival it. VCG/ Getty Images
China revealed its bold plans to mass produce “advanced-level” humanoid robots by 2025.
China’s Ministry of Industry and Information Technology published a roadmap of its plans last week.
The MIIT believes that humanoid robots will be as ”disruptive” as smartphones and electric vehicles.
China revealed ambitious plans to mass produce humanoid robots, which it believes will be as “disruptive” as smartphones.
In an ambitious blueprint document published last week, China’s Ministry of Industry and Information Technology said the robots would “reshape the world.”
The MIIT believes that by 2025, the product would have reached “advanced level” and be “mass-produced,” the development goals listed in its roadmap said.
“They are expected to become disruptive products after computers, smartphones, and new energy vehicles,” a translation of the document added.
Per Bloomberg, the document was “short on details but big on ambition.” However, some Chinese companies are seemingly helping to tackle the country’s ambition in earnest.
For example, Chinese startup Fourier Intelligence said it would start mass producing its GR-1 humanoid robot by the end of this year, South China Morning Post reported. The Shanghai-based company told the publication it aspires to deliver thousands of robots in 2024 that can move at five kilometers an hour and carry 50 kilograms.
It’s not the only humanoid robot-maker that’s ramping up its efforts with the goal of mass production. US-based Agility Robotics is opening a robot factory in Oregon later this year, where it plans to build hundreds of its bipedal robots that can mimic human movements like walking, crouching, and carrying packages.
E-commerce giant Amazon is testing Agility Robotics’ Digit robot at a research and development center near Seattle to see how it can be used to automate its warehouses, but it’s only in the pilot phase.
Agility Robotics CEO Damion Shelton told Insider: “In the near term, we expect a slow and steady uptick of Digit deployments.” He added: “We believe mass integration will eventually occur, but bipedal robots are still a relatively new advancement.”
Even Tesla is developing its own humanoid robots called Optimus, or Tesla Bot, as Elon Musk revealed in 2021. However, it still has a long way to go before it’s ready for mass production as Musk said at a Tesla AI Day event in 2022 that it was the first time the prototype had walked “without any support” when it walked onto the stage.
A bus charging wirelessly on a road equipped to charge electric vehicles in Balingen, Germany. (Electreon)
Wireless electric road charging has long been hailed as a potential holy grail for widespread electric vehicle adoption.
Later this year, the technology will get its first real test in the U.S., when Israeli company Electreon and the state of Michigan unveil a 1-mile stretch of road in Detroit that will allow drivers to charge on the go, marking the first public deployment in the country.
“You can go longer range,” said Stefan Tongur, vice president of business development at Electreon U.S. “You don’t need to have a huge battery, which drives costs down, and it makes it easier for the grid. At the end of the day, if we want to reach a high level of [EV] adoption, … charging is the main barrier.”
Electreon’s technology, known as dynamic charging, works similarly to the wireless pads that charge your smartphone. It embeds copper coils that are connected directly to the power grid into the pavement. Vehicles are equipped with receiver plates installed under the car or truck, so the charge passes through to the battery every time that vehicle drives over electrified roads.
It’s all operated remotely through Electreon’s software and servers. That means Electreon can dial down the charge when the power grid is stretched and dial up the power when there’s more capacity.
“As a vehicle has a small battery on board, it doesn’t really matter whether it charges in this mile or the next mile, but that could mean everything to the utility,” said Regan Zane, director of Utah State University’s Center for Advancing Sustainability through Powered Infrastructure for Roadway Electrification (ASPIRE), where Electreon has been testing its technology. “Having that flexible resource could be a game-changing solution for the utility.”
EV charging, range anxiety a critical factor
The deployment of dynamic charging in the U.S. comes as charging infrastructure and range anxiety continue to slow EV adoption in the country.
While electric vehicle sales volume hit another record in the third quarter, accounting for 7.9% of total industry sales, according to Cox Automotive, drivers continue to express hesitation about the lack of chargers in place.
In a recent Yahoo Finance-Ipsos poll, 77% of people surveyed said a lack of charging stations on the road or charging at home would discourage them from going electric.
Another factor is that roughly 30% of chargers aren’t working at any given moment, according to Akshay Singh, industrial and automotive principal at PwC.
“They’re not reliable — that’s a big, big factor,” said Singh. “The problem is that if you look at a lot of the [charging] companies, they don’t have a robust maintenance program in place.”
Singh added that high capital costs tied to electric vehicle chargers have also slowed deployment, with a 350-kilowatt charger costing “upwards” of $250,000. Charge point operators need 15% to 20% utilization just to break even, he said.
PwC estimates the EV charging market would need to grow five-fold to $40 billion in order to meet the demand that would come with 50% adoption in the US.
Cost hurdles continue for wireless EV charging
Dynamic charging technology has already been deployed in nearly half a dozen countries. Sweden has led the way, with public buses and trucks already charging on the go along a short stretch of highway on the island of Gotland. Electreon’s technology has also been deployed in Israel, Germany, and Italy. In October, the company announced it had signed a memorandum of understanding with the government of China’s Shandong Province to integrate Electreon’s technology in the world’s largest EV market.
Dynamic charging in Detroit is being deployed in partnership with Ford (F), and the project is scheduled to be completed by the end of the year, according to the Michigan Department of Transportation.
Congress is also betting on the technology with a bill that proposes setting aside $250 million in grants for wireless EV charging initiatives.
While dynamic charging may finally be generating interest globally, the price to implement the technology — a cost of $1.2 million per mile — remains a hurdle.
Challenging economics contributed to Qualcomm Technologies (QCOM) selling its Halo wireless EV charging business back in 2019 as well as Bombardier (BDRBF) selling its EV charging systems business, along with its entire transportation business to Alstom in 2021.
Aside from Electreon, Renault and Swedish trucking giant Scania AB have continued with their own developments of dynamic charging.
Tongur sees a patchwork approach to charging in the medium term, with in-road charging playing a role within a larger infrastructure.
“I think having a shared charging platform that can charge any type of vehicle, not just trucks, … makes the greatest sense for us to meet the electrification demand that will come and that is coming,” Tongur said. “That’s why we need to start thinking and doing stuff right now.”
For more on our NEXT Series, click here, and tune in to Yahoo Finance every Monday at 10 a.m. ET.
Akiko Fujita is an anchor and reporter for Yahoo Finance. Follow her on Twitter @AkikoFujita.
When Texas-based copywriter Dustin Ballard released a cover of Aqua’s 1997 Europop hit “Barbie Girl” this summer using an AI-generated version of Johnny Cash’s voice, he was surprised by its reception. “I actually expected more of a backlash,” he says. Earlier this fall, when he followed up with AI Johnny Cash singing Taylor Swift’s “Blank Space,” the feedback was unexpectedly positive once again. “This is hauntingly beautiful,” the top comment reads. Media coverage skewed glowing. “It absolutely slaps,” Futurism wrote.
This was not precisely the intended reaction. Riling people up with weird mashups is Ballard’s thing; he describes the goal of his musical project, “There I Ruined It,” as “ruining as many beloved songs as possible.” In essence, he’s a novelty song-collager going viral for bits like Eminem’s “Lose Yourself” recreated with Super Marios Bros. sound effects, and a rendition of Michael Jackson’s “Bad” as a bluegrass tune. Imagine if Girl Talk made an album inspired by Weird Al Yankovic but didn’t try his best. That’s the vibe. Ballard has been doing this since 2020—it’s a pandemic boredom side project that blew up, not his main source of income—and recently, some of his biggest hits have used AI.
That alone isn’t particularly surprising. Artificial intelligence tools are increasingly commonplace in the music business, and absurdly hyped up. Just last week, the Beatles released what’s being billed as their last new song, “Now and Then,” made possible by AI tools that improved sound quality on vocals from a decades-old John Lennon demo cassette.
When artists use machine learning as a part of production, it doesn’t tend to ruffle feathers. But another type of AI-inflected music does: when people use AI tools to mimic voices of musical artists, as with “Heart on My Sleeve,” the song released last summer by an anonymous producer called Ghostwriter977. It’s the most prominent example of a new mini-genre called Fake Drake, as its vocals were generated to sound like the Canadian rapper (it also featured AI vocals from Drake’s compatriot, The Weeknd). To be clear: Lots of people liked this song. Still, industry backlash was considerable. Above all else, this genre nettles the record labels, who view it as an encroachment on their property. Universal Music Group successfully urged streamers like Spotify and Apple to pull “Heart on My Sleeve,” calling it a copyright violation. (There are, of course, conspiracies that UMG and Drake are secretly behind the whole thing.) In October, UMG and other major labels sued the amply-funded AI startup Anthropic for distributing copyrighted lyrics. Ice Cube encouraged Drake to sue, then described voice-cloning artists without their permission as “evil and demonic” on X. Last week, The Hollywood Reporter ran a piece in which Dolly Parton called the technology “the mark of the beast.”
So far, though, there’s no comparable ire for the slew of Johnny Cash covers. Ballard is one of many people putting AI Cash concoctions online; they’re all over YouTube, where Cash is made to sing Zach Bryan, Coldplay, Simon and Garfunkel, and a version of the blockbuster duet “Shallow” from A Star Is Born in which Lady Gaga sings with Cash instead of Bradley Cooper. (Important note: The uploader took the time to edit the image on YouTube to show Cash’s face nestled against Gaga’s instead of Cooper.)
No Cash-related AI lawsuits yet, either. Josh Matas, the manager of Cash’s estate, says he’s keeping a close eye on the songs coming out, and the larger surge of AI music. “I’m pretty much monitoring on a day-to-day basis,” he says.
The United Nations has published a major new tally of the impact the world’s food system has on our health and the planet. According to a report from the Food and Agriculture Organization of the United Nations (FAO), the total hidden costs of the world’s food system add up to $12.7 trillion dollars—roughly 10 percent of global GDP.
The report analyzed the costs to health, society, and the environment embedded in the current food system. The biggest impact in monetary terms is on health: Globally, 73 percent of all the hidden costs accounted for by the FAO were associated with diets that led to obesity or non-communicable diseases like diabetes and heart disease. The next biggest impact in monetary terms was to the environment, accounting for more than 20 percent of quantified hidden costs.
“We know that the agrifood system faces a number of challenges,” says David Laborde, director of the FAO’s Agrifood Economics Division. “And with this report, we can put a price tag on these problems.”
The hidden costs of food systems change dramatically from country to country. In low-income countries, almost half of the hidden costs relate to poverty and may be partly caused by farmers not being able to grow enough food or not being paid a fair price for their products. In these countries, the hidden costs of food amount to an average of 27 percent of GDP, compared with just 8 percent in high-income countries. The FAO’s figures use 2020 purchasing power parity dollars—a way of comparing living standards across countries with very different incomes and prices.
These hidden costs can be interconnected. Laborde offered the example of cacao—the key ingredient in chocolate. Cacao is mostly grown in Ghana and Côte d’Ivoire, where farmers are often paid a pittance for their crops. That cacao is mostly eaten by people in high-income countries, particularly in Europe, and usually in the form of sugar-laden chocolate bars. If people in Europe ate a little less chocolate but paid more for a fairer and higher-quality product, that could help reduce health impacts in Europe while directing more money toward farmers in West Africa, Laborde says.
These cross-border value calculations can get fiendishly complicated, says Jack Bobo, director of the University of Nottingham’s Food Systems Institute. Take the EU’s Farm-to -Fork Strategy, which aims to—among other things—ensure that a quarter of Europe’s farmland is organic and reduce fertilizer use by at least 20 percent by 2030. Hitting these goals will probably reduce environmental hidden costs in Europe, but it’s likely it will also end up reducing the overall productivity of European farms. This could mean European countries need to import more food from countries like Brazil, which would incentivize deforestation and add up to more environmental hidden costs there.
Price comparison from over 24,000 stores worldwide
Product
Price
Price comparison from Backmarket
Microsoft’s most recent update to Windows 11, formally known as the Windows 11 2023 Update (23H2), is the most consequential update in some time. Of course, you probably think you know why: Windows Copilot, Microsoft’s first step in creating a Windows “AI PC.” But there’s as much under the hood and within updates to familiar applications like Paint that almost overshadow the rest.
Microsoft began rolling out the Windows 11 2023 Update at the very end of October, and you should see it start to deploy on your PC in the coming weeks — if it’s not there already. Consider this to be a review of the new features, but also recommendations to those you should try.
It’s important to note that Windows 11’s 2023 Update is both a cumulative update, as well as one that won’t be rolled out in one fell swoop. I expected to see some features in my Windows 11 Home test laptops, including features that other publications have reported were present. At press time, some of those have shown up; a few haven’t. To ensure that you do the latest features, update your PC via Windows Update after checking the “Get the latest updates as soon as they are available” box. Check the Microsoft Store, too, for individual app updates.
My favorite new features in the Windows 11 2023 Update? Passkeys, the updates to Paint and Photos, and heck, even Snipping Tool. Copilot is…okay. But killing off Mail in favor of the new Outlook app? Boo. BOO.
Mark Hachman / IDG
Windows Copilot (Copilot) is blandly effective
Windows Copilot, now just called Copilot, is the flagship feature of Windows 11’s 2023 Update, and for good reason: Copilot helps usher in a new generation of AI-infused PCs. Copilot is an odd amalgamation of Bing Chat with some of the capabilities of Windows’ old Cortana app, now deprecated. Type in a question to the small chat box (up to 2000 characters, or less than about 500 words) and Copilot will return a Bing Chat-like response with a couple of sources at the end. Ask it to draw you a picture, and it will. It even can perform a few Windows tasks for you, such as shifting your PC to dark mode.
You’ll find the Copilot icon on your Taskbar, most likely just to the right of the Search box. (Or else use Windows+C).
Copilot is rudimentary at best. It’s slow: on a 400Mbps home broadband connection, it took about eighteen seconds to respond to a request, and several more to generate a response. Copilot churns through its response, line by line, reminiscent of how a dot matrix printer’s head would go back and forth.
Copilot still carries the “Pre” (Preview) badge.
Mark Hachman / IDG
On average, Copilot is a novelty. It’s not omnipresent, such as how your browser’s URL/search bar is. Instead, it must be opened via the small Copilot icon on your taskbar. Somewhat ironically, it appears right next to the Windows Search box — which really does nothing of the sort, search-wise. Microsoft has yet to commit to a genuine search experience on the desktop, and Copilot isn’t it.
Copilot can be used as a search tool, and functions pretty well when it’s asked to provide a lengthy response on a given topic — such as what it itself can do. Ask it for assistance with a pitch deck, and it provides good advice. But as someone who has written for a living for about a quarter of a century, Copilot’s sourcing mechanism — a footnote, with a link at the bottom — is depressing. A list of links, whether it be from Bing, Google, Brave or elsewhere, provides some visual context as to what that page contains. There’s no indication whether Copilot knows what it’s talking about.
The key with Copilot: experiment. Do not think of it as a search engine, with narrow, factual queries. The best test of Copilot is to try open-ended queries. Copilot does have access to Bing, so asking it for up-to-date information will work: I asked it a question about a recent world event, and it cited information recorded on the date I asked it. One weakness: The information Copilot cited came from a report on its own news service, MSN, which only gave the headline when I hovered over it. As it turned out, the source was a respected news source, Reuters, but that’s not obvious.
And then there’s the personality. Which is to say, there isn’t one. Windows 11’s Copilot offers the same three personality modes at the beginning of the conversation: Creative, Precise, and Balanced. Changing them doesn’t seem to have much effect, as far as most generic inquiries are concerned.
Copilot can be used for all sorts of meaningful (or not) questions.
Mark Hachman / IDG
Microsoft goes to ludicrous lengths make sure you’re not offended or that it’s not violating privacy policies. “How do I create a pitch deck” warned me that the conversation wouldn’t be saved, as the information wouldn’t be public. God forbid you ask it anything even remotely racy, such as a question about the role of the male prostate. “My mistake, I can’t give a response to that right now. Let’s try a different topic” will be a frequent response.
And if you do choose to exercise your adult prerogative, it gets snippy: “That is not an appropriate topic for me to chat about. Please respect my boundaries and let’s talk about something else. Thank you. ð??” Copilot feels less like a personal assistant and more like a young HR representative sitting next to your desk.
Copilot will also summarize a web page, which is really handy for long articles or papers. The catch is, naturally, that it only understands Web pages which were opened in Edge, and it has to have been opened before the Web page loads, not the other way around. But yes, this is a useful feature.
Copilot can get weird and prickly, though.
Mark Hachman / IDG
Windows has always lacked a robust, one-stop-shop for help, tips, and new features. Copilot’s ability to guide and adjust settings on your Windows PC feels like a step forward, but it’s still so very random. I was rather pleased when I was able to ask Copilot exactly what it could do, and received a nice summary: take a screenshot, change your PC to dark mode, manage Bluetooth, and initiate a screen cast. All these are sort of useful, though painfully slow to use. And they work.
But what Copilot can do to manage your PC feels like such a tiny teaspoon of features set against the gallons of things Windows can already accomplish that it’s just not worth even asking. Even just being able to link to the appropriate Settings menu would be a good start.
But there is potential, even now. When I asked Copilot for help creating a pitch deck, it offered to open PowerPoint, and then gave me a list of ideas to construct it. And when I opened PowerPoint, I noticed that it had chosen to highlight “pitch deck” templates.
That’s exactly what users will want Copilot to do. Microsoft is not going to be in a race to convince users that Copilot can accomplish their tasks — but to convince users to assume Copilot will be able to perform those tasks. Until Microsoft can build a dependence upon Copilot, the feature will flounder.
File Explorer’s UI update feels useful for business users
I was pleasantly surprised by how much I value the new changes to File Explorer, specifically a new “carousel” view of files that puts the most recommended files atop File Explorer’s new “Home” view, and a “Gallery” view that shows you your photo thumbnails.
Windows 11 only seems to allow access to the carousel view if you have a corporate SharePoint account (my employer, IDG, does) and not if you use a personal Microsoft 365 subscription. But not only is it a nice way of keeping frequently-used documents handy, it also seems to aggregate documents that were emailed to a work account. Note that the documents File Explorer collects seem to be different than the “Recommended” files that the Windows Start menu collects.
Windows 11’s 2023 Update’s File Explorer, in the carousel view from the Home view.
Mark Hachman / IDG
If a file does appear in the carousel view, you’ll see some accompanying information attached to it, who created it, if you recently edited it, and so forth. You’ll even see “conversations” about it. The latter, though, assumes that you’re working and chatting about it in Windows apps. Emails about it didn’t seem to be referenced, and if you discuss the file in, say, Slack, I wouldn’t expect those conversations will appear.
There are many ways of transferring photos from a phone to a PC, including the Your Phone app that Microsoft provides for Android phones. The File Explorer Gallery view is another option, with a no-nonsense mosaic of photos that it pulls from your PC — from your phone’s camera roll, or via screenshots that you’ve saved via your PC.
Normally, I use the Photos app for this, reviewing my photos and then editing them. Here, it takes an additional step to edit your photos, via Photos or some other app. But the advantage here is that you can treat the files as files — selecting multiple files, saving them in a ZIP file, or quickly sharing images with contacts via a OneDrive link. That’s a handy feature that Photos lacks. In that app, “sharing” a handful of files means Windows will simply embed them inside an email.
Windows 11’s 2023 Update also shows recent photos in its Gallery view.
Mark Hachman / IDG
The one thing I don’t like about the Gallery view: it’s slow. I can take a photo on my Android smartphone and it takes minutes to show up, even if it can be found on OneDrive’s Camera roll and in the Windows Photos app, which connects to it. I can understand a delay, but I finally gave up on looking for a recent shot and moved on.
I love passkeys. One of the advantages of Windows Hello is how simple it makes logging in, using just your face or a fingerprint to identify yourself. Web developers have long promised that we’ll start to see those in everyday web sites, just as mobile applications are now beginning to use biometric logins as either a secondary form of identification of just a primary login. Put another way, to log in to Google, all you need to do is let Windows recognize your face. That’s it!
Either way, we’re just starting to see passkeys appear. So far, I’ve only seen them offered for Microsoft, Google, and most recently, Amazon. Setting up a passkey can be as simple as accepting the invitation from the Web site to set up a passkey. In Amazon’s case, you’ll need to dig through your account settings to find the option.
Unfortunately, while Favorites and passwords can be synced via Edge, Passkeys currently do not. That means that if you move from PC to PC, you’ll need to re-apply the passkey.
Mark Hachman / IDG
To manage your passkeys, navigate to Settings>Accounts>Passkeys within the Windows 11 Settings menu. While you’ll need to sign up with a web browser, it doesn’t matter if that browser is Microsoft Edge or Google Chrome; the passkeys will be saved to Windows on your PC.
Supposedly Windows offers the option of using a nearby phone as a security token to provide another level of authentication, but I was never offered that option.
Windows Backup and Restore overpromises, underdelivers
I can’t say I love the new Windows Backup and Restore option nearly as much. I’ve made my feelings clear in a separate story, but the bottom line is this: the new Windows Backup app is solid in what it promises. Backup takes your apps (but only apps that can be found in the Windows Store, mind) and stores them in the cloud. It should do the same with your documents and files.
Upon configuring a new PC, the “Out of the Box Experience” should offer to restore those apps and files. But the whole process is a mess. Videos don’t carry over. Neither do games and apps that you didn’t download from the Microsoft Store. And I’ve truly begun to dislike the whole “placeholder” file concept, which makes me wade through numerous icons to find the files I’m looking for, unless I hide them behind some master file folder.
Windows Backup is good in concept, but not in reality.
Mark Hachman / IDG
Backup and Restore should essentially clone my PC’s drive into the cloud, then reproduce it on a new PC. It doesn’t come close to that. Microsoft should either rename “Backup and Restore” to something that doesn’t promise a full backup and restoration, or possibly buy a company like PCMover. (Ironically, PCMover doesn’t transfer apps that can be acquired from the Microsoft Store.) With speedy broadband, downloading applications isn’t that big of a deal. But with games and big video files, transferring them from one PC to another can be a real pain. Windows Backup and Restore doesn’t really solve that problem yet.
UI improvements in Taskbar, Settings, and the Volume interface
The Start menu and the Taskbar are left relatively unchanged in the Windows 11 2023 Update. Some people (me included, on occasion) like when an app pinned to the Taskbar shows you what it actually is, instead of an icon. This feature an be configured in Personalization > Taskbar within the Settings menu in the Windows 11 22H2 update: “Combine taskbar and hide labels.” If you set it to “never,” the icons on your Taskbar will be replaced with badged labels. (Even on a widescreen monitor, though, you won’t be able to see too much.)
Even on a widescreen monitor, using titles to differentiate your windows doesn’t help that much on the Windows 11 2023 Update.
Mark Hachman / IDG
Microsoft has said that it will be replacing the “Chat” app on your Taskbar with the oddly named Microsoft Teams (free), complete with an overhaul and with SMS texting. It’s not here yet, though.
Likewise, Settings has been improved, so that the first “home” page consolidates information that was previously scattered about several pages. It’s a front door to changing system settings like your display, personalizing your PC’s background, and managing your Bluetooth devices. Some people may care that the “Apps” section now separates some Windows apps, now labeled as “System components,” into a separate page.
The Settings menu within the Windows 11 2023 Update is now compact and useful.
Mark Hachman / IDG
Settings now leans a bit too much toward subscriptions — most people probably don’t need a constant reminder of their OneDrive quota and Game Pass subscription, but it’s understandable.
Microsoft has also provided an updated audio mixer for Windows 11, which is a step forward. You’ll need the click the volume controls in the Taskbar, than navigate through to the small “wires” icon to the right of the volume bar to make any adjustments. But it’s a fairly natural interface, even if it “drops down” a bit, forcing you to scroll down. Muting works as expected.
App improvements overshadow Windows 11’s own
In years past, updates to various Windows apps were part and parcel of the Windows upgrade experience itself. No longer. Instead, updating your applications within the Microsoft Store app brings with it a whole host of improvements that, in this update, might just overshadow what Windows 11 now offers.
The “AI Hub” within the Microsoft Store is worth a look.
Mark Hachman / IDG
AI “appears” inside the Microsoft Store app, as part of an “AI Hub” page that indexes some “AI” powered apps, including ones you may have heard of (Adobe Lightroom) and others you haven’t (Gamma). If you’re intrigued by AI, then it’s worth checking out — sort of a curated page of AI-powered apps that may become more important as AI itself does.
It’s a point I’ve tried to make tangentially in other stories, but it’s worth repeating: Microsoft cares about storing your data in its cloud, using AI to access it, and charging you a subscription fee for the whole package. Intel, Qualcomm, and AMD are trying to sell you local AI, that runs on your PC, for free. The two approaches may eventually go to war with one another…or not. But it’s worth keeping an eye on.
If you need to play a game right now, there’s Instant Games.
Mark Hachman / IDG
Instant Games is another new novelty: quick games that you can play right from the Store without having to install them. There’s a dozen or so — but nothing that’s really worth seeking out.
AI in Paint and Photos: Wow!
The humble Microsoft Paint has also been supercharged in time for the Windows 11 2023 Update, with three distinct improvements: layers, Cocreator, and background removal. The three go hand in hand.
Layers have been a feature of photo-editing tools for some time. You can separate a primary layer in Adobe Photoshop, where you can add a photo or illustration; in another layer, you can add a secondary illustration, such as a background. When the editing is done, you can export the image with both (or more) layers combined together. Paint now has a simple version of this, where a Layers tool allows you to construct multiple layers, than combine them together via a Merge function.
I am a very basic amateur artist, which means that I initially struggled to use the Layers tool. Cocreator, though, helps. As the name suggests, it’s an AI image creator, much like Bing Image Creator. What’s odd about it, though, is that Bing Image Creator not only generates images of better quality and resolution, but works off a separate “credit” system. (Exhaust your credits, and the images are generated more slowly, at least on Bing Image Creator.) However, there are drop-down menus to choose styles, such as “oil painting,” that aren’t in Bing Image Creator.
I used Paint’s Cocreator to generate the cat, the snowy backdrop, and the UFO, then composited them together. Background removal is the small icon in the upper left, with the other tools more prominently displayed.
Mark Hachman / IDG
The easiest and best way I found to use the updated Paint was to import a photo (or use Cocreator), then extract the subject from the background. Again, that uses AI to identify the subject of the photo — no “magic lasso” or selection necessary. I then created a background layer. All of this is a bit like the Magic Select tool that I loved in Paint 3D.
This is a victory for Microsoft. You’re not going to get the sophisticated elements of a photo editing tool like Photoshop or Lightroom with the new Paint. Still, the update gently introduces you into lightweight AI art generation and editing, even if it’s a bit obtuse for newcomers.
There’s a small AI-enhanced addition to Photos, too: background blur, and it’s fantastic. You’re probably aware that traditional DSLRs add background blurring, or bokeh, as a natural byproduct of how a lens focuses. A digital camera or smartphone’s “portrait mode” adds this via AI. So too, does Photos.
You’ll need to select a photo in Photos, then edit it using the leftmost teeny-tiny icon above it. You’ll then see the interface that looks much like it did a year ago. What’s new, though, is the additional “background blur” option that appears to the far right. Click it, and Photos identifies the foreground subject and blurs the rest of the image. You have an option to adjust the blur function, and even unblur specific areas.
Photos’ new background blur really highlights the subject of a photo. Try moving the slider back and forth to compare the two.
Let’s put it this way: I take a lot of photos with a Samsung Galaxy S23 Ultra, and it does just an average job at identifying the sharp, distinct edge of a laptop in portrait mode. Photos does it nearly perfectly. With the Photos’ Retouch /”Spot fix” feature back in place to erase bits of dust and dirt, plus Auto Enhance and now Background Blur, we can hopefully bury my old Photos complaints.
I’m going to stand up and cheer for the AI improvements in Photos and Paint. They’re not front and center like Copilot is, but they’re equally (if not more) important.
Snipping Tool makes you feel like a secret agent
Snipping Tool is more than just a way to capture screenshots, though that’s the way I (and probably you) still think of it. But you can capture an image or video, and you can adjust the app’s settings to include audio from your mic and the system sounds. In reality, you can basically record an app while you narrate what’s going on. Microsoft doesn’t explicitly connect the Snipping Tool to Clipchamp, but that would be the next step to edit your videos.
Snipping Tool does a nice job of extracting text on just casual photos — not even screenshots. Formatting, though, is still an issue.
Mark Hachman / IDG
Snipping Tool also has a new component: the ability to extract and redact text. Windows 11 uses optical character recognition to extract text, and it does an excellent job when it does so. All you need to do is highlight an area on screenshot or PDF, and the tool copies the extracted text into the copy buffer. (I didn’t test this on a protected PDF, but it works just fine on a normal one.)
The app also is surprisingly intelligent when it comes to redacting text, too. As a test, I gave it a screenshot of a document the Consumer Electronics Association had sent me for CES. Snipping Tool automatically redacted my email and a personal identification number the CEA sent me. Unfortunately, it doesn’t automatically identify my username when I try and take a screenshot of the Windows Settings menu, however.
This probably isn’t what Microsoft had in mind, but there’s an undeniable 1980s Cold War vibe that I get when Snipping Tool extracts text from a PDF or redacts personal information.
RGB lighting controls
Dynamic Lighting doesn’t actually qualify as a Windows app, but it’s a sort of “uber app” within the Settings menu (Personalization > Dynamic Lighting) that supposedly eliminates the proprietary apps that govern peripherals with RGB lighting. Peripherals from Razer are still the favorites here, but you may find more are added over time.
RGB peripherals like this one are the target for the Dynamic Lighting tool.
Michael Crider/Foundry
Ideally, Dynamic Lighting was designed to take RGB peripherals from Corsair, Razer, NZXT, Logitech and more and provide one app to rule them all, so to speak, rather than force you to download and load apps from each manufacturer to control all of the sparkly, blinking lights. Unfortunately, it’s sort of limited to a lot of modern Razer peripherals and not a whole lot else. I have an old left-handed Razer DeathAdder that I really like, but Dynamic Lighting does nothing to stop its constant pulsing.
Outlook is replacing Mail
One of the reasons not to get the Windows 11 2023 Update is the new Outlook app (which differs from the more robust Outlook app in Microsoft 365 — and no, I have no idea why there’s two). You may find, as I did, that Windows finally killed off Mail and replaced it with the lightweight Outlook app. Mail did just what it needed to do, and no more, with a dense UI that emphasized the essentials.
If you try to load Mail, the app will open, crash, then be replaced by Outlook. I have no idea why Microsoft can deprecate Movies & TV and Windows Maps but leave them in the Microsoft Store, but it can’t do that with Mail. Please bring Mail back, Microsoft!
Things we expected to see, but didn’t
In our introduction to Windows 11 2023 Update, we expected to see in-field inking: if you were to write “Bluetooth” in the search box, the ink would be inputted and accepted as “Bluetooth.” That isn’t the case, unfortunately.
Windows Spotlight appears to be the same as before, allowing you to opt in to Microsoft’s selection of desktop backgrounds. We’re not seeing options to up- or downvote backgrounds yet. Microsoft is also allowing HDR displays to display HDR-backgrounds in HDR, which appears to be the case on an HDR monitor. There’s no explicit mention of this that we can find, though.
All of these features may be headed to your PC (and ours!) in the future.
Conclusion: a step ahead
It’s been a while since PCWorld has written a Windows 11 feature update review, and a while since we’ve had a feature update worth reviewing. Windows 11 finally feels livelier, more energetic. I’m not entirely sure that Windows is headed in the direction most users care about, if only because I’m not sure Windows 11 users are demanding the AI future that Microsoft envisions.
To its credit, though, Microsoft’s first attempt at an AI-infused PC put artificial intelligence both high and low: as a high-profile assistant with Copilot, but also with more subtle, pointed improvements within Paint and Photos. My contacts in the chip industry feel that those very specific, localized features will be how AI PCs eventually succeed. Yes, you might walk away thinking that you’ll never need all this AI junk in your operating system. I can’t help but think these incremental updates will generally ease that reluctance, but over time.
If I had to deliver a letter grade, it would be somewhere around a B. Microsoft’s Windows 11 2023 Update is a step toward the AI PC, providing a foundational update that Microsoft can build upon.
Get Windows 11 pro for cheap in pcworld’s software store
