An unseen, non-human hand moving the cursor across your computer screen and typing without using the keyboard in fiction is usually a sign of malicious AI hijacking something (or a friendly ghost helping you solve mysteries like the TV show Ghost Writer). Thanks to Anthropic's new computer use feature for its AI assistant Claude, there's a much more benevolent explanation now.
Fueled by an upgraded version of the Claude 3.5 Sonnet model, this AI – dubbed 'computer use' – lets you interact with your computer much like you would. It takes the AI assistant concept a step beyond text and a voice, with virtual hands typing, clicking, and otherwise manipulating your computer.
Anthropic bills computer use as a way for Claude to handle tedious tasks. It can help you fill out a form, search and organize information on your hard drive, and move information around. While OpenAI, Microsoft, and other developers have demonstrated similar ideas, Anthropic is the first to have a public feature, though it's still in beta.
"With computer use, we're trying something fundamentally new," Anthropic explained in a blog post. Instead of making specific tools to help Claude complete individual tasks, we're teaching it general computer skills—allowing it to use a wide range of standard tools and software programs designed for people."
The computer use feature is due to Claude 3.5 Sonnet's improved performance, particularly with digital tools and coding software. Though somewhat overshadowed by the spectacle of the computer use feature, Anthropic also debuted a new model called Claude 3.5 Haiku, a more advanced version of the lower-cost Anthropic model, though once capable of matching Anthropic's previous highest performing model, Claude 3 Opus, while still being much cheaper.
Invisible AI assistanceYou can't just give an order and walk away, either. Claude's control of your computer has some technical troubles as well as deliberate constraints. On the technical side, Anthropic admitted Claude struggles with scrolling and zooming around a screen. That's because the AI interprets what's on your screen as a collection of screenshots, and then it tries to piece them together like a movie reel. Anything that happens too quickly or that changes perspective on the screen can flummox it. Still, Claude can do quite a lot by manipulating your computer, as seen above.
Unrestrained automation has obvious perils even when working perfectly, as so many sci-fi movies and books have explored. Claude isn't Skynet, but Anthropic has placed restraints on the AI for more prosaic reasons. For instance, there are guardrails stopping Claude from interacting with social media or any government websites. Registering domain names or posting content is not allowed without human control.
"Because computer use may provide a new vector for more familiar threats such as spam, misinformation, or fraud, we're taking a proactive approach to promote its safe deployment. We've developed new classifiers that can identify when computer use is being used and whether harm is occurring," Anthropic wrote. "Learning from the initial deployments of this technology, which is still in its earliest stages, will help us better understand both the potential and the implications of increasingly capable AI systems."
You Might Also LikeStriking machinists voted to reject an agreement that would have boosted wages by 35%. It’s another blow for Boeing, which reported a $6 billion quarterly loss on Wednesday.
(Image credit: Mario Tama)
Both candidates are in the crucial swing state this week. Trump had two events on Wednesday, while Harris, ahead of her own visit, answered "Yes" at a CNN town hall when asked if Trump is a fascist.
(Image credit: Anna Moneymaker)
I felt like I opened a can of worms when I asked Samsung about its stance on AI photo editing at Galaxy Unpacked back in January. “There is no such thing as a real picture,” Patrick Chomet, Samsung’s Head of Customer Experience, told me at the time – a nihilism-tinged soundbite that, in fairness, came as part of a nuanced and perfectly valid philosophical commentary on the nature of photography (in the same interview, Chomet rightly talked up the importance of validating authenticity).
Marques Brownlee recently included Chomet’s quote in an X post highlighting the differing opinions of Samsung, Apple, and Google on this very subject. Still, ultimately, all of the major phone manufacturers are offering similar editing features (Magic Eraser, Clean Up, AI Eraser, and so on).
Qualcomm, for its part, acknowledges that AI photo editing is an exciting new frontier but also one fraught with awkward ethical implications. During a roundtable interview at Snapdragon Summit, the company’s Mobile Handset SVP, Chris Patrick, explained why he believes that AI photo editing “is not as simple as a real image and fake image.”
LOL "What is a photo?"Samsung: "Actually, there is no such thing as a real picture"Google: "authentic to your memory and to the greater context, but maybe isn’t authentic to a particular millisecond"Apple: "a personal celebration of something that really, actually happened" https://t.co/mwy3Yb9KaYSeptember 23, 2024
“It’s an interesting question, because the human brain is not digital,” the former engineer explained. “When you perceive the sunset outside, you don't perceive the sunset absent of context. You know where you are. You know what you're smelling. You know what you're hearing, You know what you're feeling. You know what a sunset looks like, right? All of that is part of how you perceive the space."
“So then, when we capture an image, should it just be the raw response from the sensor? Or is it right to include context in how that picture is created? I'm not an expert. I don’t have a PhD in this topic, but I do think it's not as simple as a real image and fake image. I think context matters, and extracting the very best we can from the whole situation is an accurate reflection of what the eye and the brain do as well."
When you perceive the sunset outside, you don't perceive the sunset absent of context.
Chris Patrick“[However],” Patrick continued, “if there's a picture of Rui [Guo, Honor CMO, also in attendance at the roundtable] and I on the moon, it should be very clear that that is not, in fact, a picture of Rui and I on the moon. That’s not correct – that’s manipulated.
“So, for us, it’s very clear that there has to be some mechanism to distinguish the other extreme – those manipulated images – for people. We’re building incredible technology together but we want to make sure that, in the end, it does not have a destructive impact, where people can no longer tell the difference between truth and fiction.”
More AI is on the way Honor CMO Ray Guo announcing the Honor Magic 7 Pro on stage at Snapdragon Summit (Image credit: Honor)Honor CMO Rui Guo also attended the Snapdragon Summit roundtable, having just announced that the Honor Magic 7 Pro – slated for release in Europe next year – will launch with the industry's first on-device generative AI portrait enhancement. So, is the company using AI to make portraits appear better than they actually are? Or is the aim to mimic real-life details that Honor’s hardware can't capture on its own?
“On the one hand,” Guo explained, “we have to make sure that the photo itself is good by modern consumer standards. Everyone knows what [constitutes] a good and bad photo. But on the other hand, we definitely want to retain the authenticity of the photo. We're not trying to Photoshop the photo for the consumer, right? We should give the authority to the users at end of the day.”
And that, I think, is the crux of the matter. Tools like Magic Eraser and Clean Up become ethically dubious when users – or, more worryingly, politicians – attempt to pass tampered images off as real, but as a means of occasionally removing strangers from family vacation photos, AI photo editing is, for many people, undeniably useful.
Neither Qualcomm nor Honor is forcing these tools onto their consumers – if you want to take a photo in the same way as you always have done, you still can. Rather, companies are giving users the freedom to choose how to capture moments that matter to them. Sure, it’s a can of worms, but as Qualcomm’s Chris Patrick noted, “it’s not as simple as real image and fake image.”
You might also likeStrands is the NYT's latest word game after the likes of Wordle, Spelling Bee and Connections – and it's great fun. It can be difficult, though, so read on for my Strands hints.
Want more word-based fun? Then check out my Wordle today, NYT Connections today and Quordle today pages for hints and answers for those games.
SPOILER WARNING: Information about NYT Strands today is below, so don't read on if you don't want to know the answers.
NYT Strands today (game #235) - hint #1 - today's theme What is the theme of today's NYT Strands?• Today's NYT Strands theme is… My cup of tea
NYT Strands today (game #235) - hint #2 - clue wordsPlay any of these words to unlock the in-game hints system.
• 'Hot' drink
NYT Strands today (game #235) - hint #4 - spangram position What are two sides of the board that today's spangram touches?First: top, 3rd column
Last: bottom, 4th column
Right, the answers are below, so DO NOT SCROLL ANY FURTHER IF YOU DON'T WANT TO SEE THEM.
NYT Strands today (game #235) - the answers (Image credit: New York Times)The answers to today's Strands, game #235, are…
I realize that this confession might get me kicked out of my home nation, the UK, but I have never drunk tea. Not a single cup has ever passed my lips – which meant I was at an ever greater disadvantage today than some will be.
Even setting aside my own drinking habits, this Strands puzzle is strangely specific. Why MASALACHAI? We've previously had one Strands game about coffee, with the likes of MOCHA and LATTE, but even though I don't drink that either I was at least able to make informed guesses based on just having lived in a world in which other people regularly do drink it. But the theme here isn't 'types of tea' – with ginseng or green or breakfast or whatever – it's specifically the ingredients for masala chai. That may well be very popular, but if you don't drink it, you're a bit stuck.
Fortunately I only needed one hint. I spotted PEPPER by accident, then asked Strands to give me a clue and got ANISE. Though I still had no idea what the spangram might be, I had a strong hunch that the other answers would be spices, so found the likes of CARDAMOM and NUTMEG by trial and error. That left the spangram, which revealed itself on the board to make up for my lack of knowledge in this subject.
How did you do today? Send me an email and let me know.
Yesterday's NYT Strands answers (Wednesday, 23 October, game #234)Strands is the NYT's new word game, following Wordle and Connections. It's now out of beta so is a fully fledged member of the NYT's games stable and can be played on the NYT Games site on desktop or mobile.
I've got a full guide to how to play NYT Strands, complete with tips for solving it, so check that out if you're struggling to beat it each day.
Good morning! Let's play Connections, the NYT's clever word game that challenges you to group answers in various categories. It can be tough, so read on if you need clues.
What should you do once you've finished? Why, play some more word games of course. I've also got daily Wordle hints and answers, Strands hints and answers and Quordle hints and answers articles if you need help for those too.
SPOILER WARNING: Information about NYT Connections today is below, so don't read on if you don't want to know the answers.
NYT Connections today (game #501) - today's words (Image credit: New York Times)Today's NYT Connections words are…
What are some clues for today's NYT Connections groups?
Need more clues?
We're firmly in spoiler territory now, but read on if you want to know what the four theme answers are for today's NYT Connections puzzles…
NYT Connections today (game #501) - hint #2 - group answersWhat are the answers for today's NYT Connections groups?
Right, the answers are below, so DO NOT SCROLL ANY FURTHER IF YOU DON'T WANT TO SEE THEM.
NYT Connections today (game #501) - the answers (Image credit: New York Times)The answers to today's Connections, game #501, are…
I wonder how many Connections players also play Wordle? I'd guess that the crossover is pretty high – in fact, I'd wager that almost everyone who plays this game also tries to solve that one each day too. And with that in mind, I wonder how many made the mistake that I did today and played four Wordle start words as a group.
It's hardly surprising that people would be thrown off the scent here. ADIEU, CRANE, SLATE and AUDIO absolutely make sense as a group, because they are the two most popular start words and the two best (according to WordleBot, the NYT's in-game helper tool). But no, this was some classic NYT misdirection. Instead, SLATE went with BILL, PROGRAM and SCHEDULE to make LINEUP, the easiest group, while AUDIO went with COOKING, GAMES and NEWS to make NYT OFFERINGS (green). At least that one was still NYT-related, eh?
With those out of the way it was all a lot easier, and I spotted the French words group of ADIEU, BELLE, PAIN and TEMPS to complete the set without the need to solve blue. That one was THINGS MADE BY FOLDING PAPER, which I honestly don't think I would have got on my own.
How did you do today? Send me an email and let me know.
Yesterday's NYT Connections answers (Wednesday, 23 October, game #500)NYT Connections is one of several increasingly popular word games made by the New York Times. It challenges you to find groups of four items that share something in common, and each group has a different difficulty level: green is easy, yellow a little harder, blue often quite tough and purple usually very difficult.
On the plus side, you don't technically need to solve the final one, as you'll be able to answer that one by a process of elimination. What's more, you can make up to four mistakes, which gives you a little bit of breathing room.
It's a little more involved than something like Wordle, however, and there are plenty of opportunities for the game to trip you up with tricks. For instance, watch out for homophones and other word games that could disguise the answers.
It's playable for free via the NYT Games site on desktop or mobile.
Quordle was one of the original Wordle alternatives and is still going strong now more than 1,000 games later. It offers a genuine challenge, though, so read on if you need some Quordle hints today – or scroll down further for the answers.
Enjoy playing word games? You can also check out my Wordle today, NYT Connections today and NYT Strands today pages for hints and answers for those puzzles.
SPOILER WARNING: Information about Quordle today is below, so don't read on if you don't want to know the answers.
Quordle today (game #1004) - hint #1 - Vowels How many different vowels are in Quordle today?• The number of different vowels in Quordle today is 4*.
* Note that by vowel we mean the five standard vowels (A, E, I, O, U), not Y (which is sometimes counted as a vowel too).
Quordle today (game #1004) - hint #2 - repeated letters Do any of today's Quordle answers contain repeated letters?• The number of Quordle answers containing a repeated letter today is 2.
Quordle today (game #1004) - hint #3 - uncommon letters Do the letters Q, Z, X or J appear in Quordle today?• No. None of Q, Z, X or J appear among today's Quordle answers.
Quordle today (game #1004) - hint #4 - starting letters (1) Do any of today's Quordle puzzles start with the same letter?• The number of today's Quordle answers starting with the same letter is 0.
If you just want to know the answers at this stage, simply scroll down. If you're not ready yet then here's one more clue to make things a lot easier:
Quordle today (game #1004) - hint #5 - starting letters (2) What letters do today's Quordle answers start with?• D
• R
• C
• L
Right, the answers are below, so DO NOT SCROLL ANY FURTHER IF YOU DON'T WANT TO SEE THEM.
Quordle today (game #1004) - the answers (Image credit: Merriam-Webster)The answers to today's Quordle, game #1004, are…
This was a remarkably straightforward Quordle for me, despite the repeated letters in both RALLY and LABEL. All four words are very common, as are all 10 letters that are used here. And it's that final point that probably made the most difference today; most Quordles use a far greater number of distinct characters, for instance yesterday's had 12 and Tuesday's was at 14. Nine out of the 10 were in my three start words, too, with B the only letter I had to add to the mix. With three of the answers ending in a Y, two of them ending in LY and five Ls overall, if you simply played by probability today you'll have been fine.
How did you do today? Send me an email and let me know.
Daily Sequence today (game #1004) - the answers (Image credit: Merriam-Webster)The answers to today's Quordle Daily Sequence, game #1004, are…
Hurricane Helene left destruction in its wake in western North Carolina. But elections officials are also moving heaven and earth to bring back a sense of normalcy, one vote at a time.
(Image credit: Allison Joyce)
Fujitsu has launched the FMV Zero notebook, which at only 634g (1.4 lbs), looks to claim the title of the world’s lightest 14-inch laptop.
This ultra-light device is part of FCCL’s FMV LIFEBOOK UH series and features Intel’s latest Core Ultra processors with integrated AI Boost NPUs for hardware acceleration of AI services such as Copilot on Windows 11.
The WU5/J3 is powered by Intel Core Ultra 5 125U or Core Ultra 7 155U processors and weighs between 634g and 639g, depending on the configuration. For comparison, the non-touch screen version of the Dell XPS 13 weighs 1.22 kg (2.69 lbs), approximately 92.43% heavier than the FMV Zero.
Short battery lifeThe FMV Zero is currently only available to buy in Japan through Fujitsu’s shopping site Web Mart, and comes in two models: WU5/J3 and WU4/J3. The WU5/J3 will be available in mid-November 2024 for 197,900 yen (approximately $1,320 USD), and the WU4/J3 will be released in late October 2024 for 193,300 yen (around $1,290 USD). Both models offer a range of configurations, with memory options ranging from 8GB to 64GB and SSD storage capacities from 256GB to 2TB.
Fujitsu says it achieved this lightweight design by optimizing the LCD panel, battery, and palm rest materials, as well as redesigning the internal components to reduce their weight by a few grams here and there. While the end result is a super-light laptop, the quest to trim as much weight as possible has massively impacted battery life. If you’re watching video or performing intensive tasks, you’ll run out juice in under six hours, although it can manage up to 15 hours in idle mode.
The WU4/J3, slightly heavier at 888g, offers a longer battery life of up to 11.5 hours of video playback or 30 hours of idle time. It is powered by Intel Core Ultra 5 125H or Core Ultra 7 155H processors.
The FMV Zero is equipped with AI-powered tools like noise cancellation, Umore makeup, and Quick Capture, with enhanced speaker recognition features.
More from TechRadar Pro