Agregátor RSS

** Nález kostry z doby římské nevzbudil před půlstoletím žádný rozruch ** Stáří jednotlivých kostí a analýzy DNA ale odhalily, že jde o skládačku ** V hrobu jsou kosti nejméně osmi lidí, které od sebe dělila skoro tři tisíciletí

Bitdefender has released a decryptor for the 'ShrinkLocker' ransomware strain, which uses Windows' built-in BitLocker drive encryption tool to lock victim's files. [...]

Přesně o půlnoci z 18. na 19. listopadu 2024 bude spuštěn prodej bílé limitované edice zařízení Steam Deck OLED. Uvnitř Steam Decku běží SteamOS 3 založený na distribuci Arch Linux.

Krátce poté, co Američané zvolili Donalda Trumpa prezidentem, již budoucí hlava státu rozděluje ministerstva a jiné klíčové pozice. Jednu by měl získat i jeho velký podporovatel Elon Musk. On a podnikatel Vivek Ramaswamy by měli společně vést nové „ministerstvo pro vládní efektivitu“. Úřad ...

Romanian cybersecurity company Bitdefender has released a free decryptor to help victims recover data encrypted using the ShrinkLocker ransomware. The decryptor is the result of a comprehensive analysis of ShrinkLocker's inner workings, allowing the researchers to discover a "specific window of opportunity for data recovery immediately after the removal of protectors from BitLocker-encrypted

Romanian cybersecurity company Bitdefender has released a free decryptor to help victims recover data encrypted using the ShrinkLocker ransomware. The decryptor is the result of a comprehensive analysis of ShrinkLocker's inner workings, allowing the researchers to discover a "specific window of opportunity for data recovery immediately after the removal of protectors from BitLocker-encrypted Ravie Lakshmananhttp://www.blogger.com/profile/[email protected]

AirPody Pro 2 podporují funkci sluchadla na úrovni klinických pomůcek pro osoby s vnímaným mírným až středním poškozením sluchu. Funkce je ale geoblokována (i v Česku). Rithwik Jayasimh z Indie se rozepsal o tom, jak pomocí Faradayovy klece geoblokaci obešel (𝕏).

Horizon v kostkovaném provedení je komornější, nenuceně vtipný a otevírá oblíbený svět nové sortě hráčů. V rámci Lego her na pohled vypadá parádně, ale náplní se těm nejlepším rovnat nemůže.

Od 1. března příštího roku změní sdružení CZ.NIC velkoobchodní cenu domény .CZ, a to na 160 korun (bez DPH) za jeden kalendářní rok. Dosud byla velkoobchodní cena české národní domény 145 korun (bez DPH), jedná se tedy o zvýšení o 15 korun. K této změně přistoupí sdružení CZ.NIC po více než šesti letech.

V prvních šesti zemích zdraží rodinná i individuální předplatná Microsoft 365. Firma k nepříjemnému kroku přistupuje poté, co do své kancelářské sady integrovala prvky umělé inteligence. V Austrálii, Malajsii, Singapuru, Thajsku, na Novém Zélandu a Tchaj-wanu totiž součástí Microsoftu 365 jsou ...

** Celodenní baterie vypadá na papíře relativně dobře... **...v praxi je to ale stále dost málo ** Co dělat, když budete s Apple Watch déle bez přístupu k nabíječce?

The rise of SaaS and cloud-based work environments has fundamentally altered the cyber risk landscape. With more than 90% of organizational network traffic flowing through browsers and web applications, companies are facing new and serious cybersecurity threats. These include phishing attacks, data leakage, and malicious extensions. As a result, the browser also becomes a vulnerability that

The rise of SaaS and cloud-based work environments has fundamentally altered the cyber risk landscape. With more than 90% of organizational network traffic flowing through browsers and web applications, companies are facing new and serious cybersecurity threats. These include phishing attacks, data leakage, and malicious extensions. As a result, the browser also becomes a vulnerability that The Hacker Newshttp://www.blogger.com/profile/[email protected]

With last week’s US elections, everything changed for the tech industry, especially for Microsoft. Given President-elect Donald J. Trump’s penchant for incendiary statements, off-the-cuff decisions, plans to impose massive tariffs, and willingness to use the power of his office for personal vendettas, almost anything could happen.

But based on Trump’s past actions, who he turns to for tech and economic advice, and what he’s said on the campaign trail, there’s a lot we know about how Microsoft will likely be affected by his upcoming four-year reign. 

There’s so much, in fact, that it’s more than can be covered in one column. Here, I’ll look at how Trump’s likely plans for AI will affect the company. In Part 2, I’ll look at tariffs, antitrust and climate change issues, and how Trump’s actions often are based more on personal grievances than policy.

Here’s how the president-elect’s likely AI plans could help or hurt the company.

Letting AI run free

Microsoft is the world leader in AI and has built generative AI (genAI) copilots into its entire product line. That was just the start. The company continues to invest billions and has bet its future on the technology. As genAI goes, so goes Microsoft. That means any actions Trump takes could dramatically affect Microsoft’s present and future. 

Trump has said plenty about tech through the years, but surprisingly little about AI. Based on his big tech backers, general outlook on technology, and the few things he’s said about it, we can get some sense of what he’ll do.

One of his biggest tech boosters is venture capitalist Marc Andreessen, who is opposed to any regulation of AI, and believes AI development should be absolutely unfettered. Elon Musk wields even more influence with Trump than Andreessen, though, and has become his most trusted tech adviser. Musk is generally against any government regulation over tech, but when it comes to AI, he’s a bit more nuanced.

Musk has often spoken about his fears that unchecked AI could lead to the destruction of the human race. He also supported a California bill that would have required AI to undergo safety testing before being deployed. As he wrote on X: ‘For over 20 years, I have been an advocate for AI regulation, just as we regulate any product/technology that is a potential risk to the public.” 

The bill was passed by the California legislature, but Gov. Gavin Newsom vetoed it.

That makes it sound as if Musk is all in on regulation. But that’s not really the case. His primary worry is that AI might eventually represent an existential threat to mankind — and that’s all he wants to regulate. He’s against regulations around existing real-world issues  such as intellectual property theft, or AI’s use to violate civil rights, its role in misinformation, privacy violations, its effect on jobs and more.

What this all means is that it’s unlikely AI will face much regulation under Trump. The first thing he’ll probably do is rescind Biden’s well-thought-out AI executive order that addresses everything from safety and security measures to issues related to bias and civil rights, and oversight over how genAI is produced. Because it was only an executive order, it didn’t carry the full force of law. But it did have some effect. And it was an ideal roadmap for how Congress could act on AI regulation if it wanted to.

Samuel Hammond, a senior economist at the Foundation for American Innovation, was blunt about what would happen to that order. He told Vox, “There will likely be a day one repeal of the Biden executive order on AI.”

Trump will also almost certainly want to boost AI to make sure the US leads China in it.  He told the right-wing influencer Logan Paul in an interview, “We have to be at the forefront [of AI]. It’s going to happen. And if it’s going to happen, we have to take the lead over China.”

Finally, in July The Washington Post reported that Trump’s tech boosters “are drafting a sweeping AI executive order that would launch a series of ‘Manhattan Projects’ to develop military technology and immediately review ‘unnecessary and burdensome regulations.’

How this affects Microsoft

What does all this mean for Microsoft? Under Trump it’s going to be full speed ahead on AI. Any regulations are likely to cover only the existential danger AI could pose to humanity, and not any of the many dangers it currently poses. That means the company will be free to develop AI in pretty much any way it wants.

There’s more good news for Microsoft than just being free of regulations. Because it is the world’s most powerful AI company, the Trump administration will almost certainly turn to Microsoft for help in its fight against China — and pay the company well for it. The Biden administration has already done so, much to Microsoft’s benefit. The administration brokered a deal in which Microsoft invested $1.5 billion in a powerful genAI company based in the United Arab Emirates, which then cut its ties to China and aligned with the US. 

In addition, there will be direct government AI contracts. Expect Trump to pour billions into the military use of AI. Microsoft has had contracts with the US military for decades. In just the past few years, it’s gotten a $22 billion contract to provide the US Army with 120,000 AR headsets and billions for a Joint Warfighting Cloud Capability (JWCC) contract to establish what the company calls “an enterprise-level tactical cloud.”

A new generation of AI-related military contracts under Trump will fatten the company’s bottom line even more.

The Trump wild card

All this requires that Trump acts rationally, though, which rarely occurs. If Microsoft wants those government contracts and wants to be free from regulation, it’s going to have to be obsequious to Trump, praise him when necessary and stay away from even the slightest hint of criticism. It’s going to have to close its eyes to the worst of his actions and behavior. It’s also going to have to do the same with Musk, who is as mercurial and thin-skinned as Trump. 

If Microsoft does all that, its profits will be sky high. As for what price in self-respect and its sense of itself as a moral company, only Microsoft CEO Satya Nadella and those who work there know.

Let’s be honest: Many Windows PCs don’t have great webcams. The webcam is often an afterthought where manufacturers cut costs when putting together laptops. And, if you have a desktop PC, you might not even have a webcam at all unless you go out and buy one.

But you almost certainly have multiple high-quality cameras built right into your smartphone of choice, whether you use an Android phone or an iPhone. And with the right bit of relatively simple setup, your smartphone’s high-end camera hardware can actually function as your PC’s webcam, too.

It might be just the secret to getting better video quality in your online meetings and other video calls — no extra expenses required.

Want to stay on top of the latest Windows PC features? My free Windows Intelligence newsletter delivers all the best Windows tips straight to your inbox. Plus, you’ll get free in-depth Windows Field Guides as a special welcome bonus!

Use an Android phone as a webcam on Windows 11 (wirelessly)

Up first: If you have an Android phone and a Windows 11 PC, Microsoft now offers a built-in way to turn your phone into a camera. It all happens wirelessly, so you don’t even need a USB cable. However, this does require Windows 11 — Microsoft didn’t add the new feature to Windows 10.

To set this up, open the Settings app on Windows 11, select “Bluetooth & devices,” and click “Mobile devices.” Activate the “Allow this PC to access your mobile devices” option if it isn’t already enabled. Then, click the “Manage devices” button.

From here, add your Android phone if it isn’t already connected to your PC. This will involve installing the Link to Windows app on your phone and signing in with the same Microsoft account you use on your PC.

Once everything is set up, ensure the “Use as a connected camera” option is activated.

If you have any trouble, try toggling the “Enabled” switch here to turn the connection off and back on again.

Chris Hoffman, IDG

Now, your Android phone will appear as a webcam in apps. (Want to test this? Try opening the “Camera” app built into Windows.)

When you select it as a webcam, you’ll see a notification on your Android phone. Tap it to allow the connection. You can then use the app on your phone or the floating panel on your PC to change settings.

You’ll see a floating window where you can switch between your phone’s front and back camera while using it as a camera.

Chris Hoffman, IDG

Turn a Pixel phone into a Windows webcam via USB

Do you have a Pixel phone? Google has a very convenient built-in way for your phone to function as a webcam — no extra apps necessary. Here’s what you’ll need:

• A Pixel 6, Pixel 7, Pixel 8, Pixel 9, or newer phone.
• A Windows 10 or Windows 11 PC.
• A USB cable to connect your phone to your PC.

To get started with this, plug your phone into your Windows PC with a USB cable — as if you were going to do an Android file transfer between your phone and the PC. You’ll see an Android system notification talking about USB connection settings on your phone. Tap it and then tap “Webcam” under “Use USB for.”

Your Pixel phone will then appear as a webcam to your Windows PC. You can select it as you’d select any other webcam device in your video-conferencing application of choice.

Pixel phones can easily function as USB webcams.

Chris Hoffman, IDG

Set up DroidCam for iPhone or Android

You can turn to a third-party app that’ll allow your phone to double as a completely wireless Windows webcam. There are a variety of paid applications for this, but DroidCam stands out from the pack.

Despite the name, this app works with both Android phones and iPhones! And it’s completely free at standard resolution. (You can get a “Pro” upgrade for a one-time $15 payment to enable higher-resolution video streaming.) And there’s also a watermark unless you pay the fee. But the price is still a bargain compared to competing applications that charge higher prices or even ongoing subscription fees. As a useful professional tool, it’s very reasonable.

To set up DroidCam, you’ll need to install the DroidCam app on your phone — get it from Google Play for Android or the App Store for iPhone. Then install the DroidCam client app on your Windows PC. Launch the client app from the Start menu after it’s installed and follow the instructions to link the phone and PC apps.

Here’s another option: Reincubate Camo has a lot of good reviews, but you’re looking at a $50 per year subscription for all the features rather than a one-time $15 payment.

DroidCam works with both iPhones and Android phones — as long as you have a Windows PC or Linux system.

Chris Hoffman, IDG

Try a phone manufacturer-specific Android app

While Android phones from other manufacturers may not offer the convenient webcam-over-USB feature Google offers on its Pixel phones, they sometimes do have their own solutions.

Samsung, for example, offers a “camera sharing” feature for Galaxy phones — but it only works with specific laptops also made by Samsung. According to Samsung’s website, you can only use the Galaxy camera sharing feature if you have a Galaxy Book5 Pro 360 Windows laptop from Samsung.

If you have a Motorola phone, it might support Motorola’s “Smart Connect” platform. If so, you can install Lenovo’s Smart Connect app (Lenovo owns Motorola) and use it to position your Motorola phone as a webcam from your PC.

Overall, you’re generally better off going with the more broadly applicable solutions, such as the ones I mentioned. But if your phone has a built-in option provided by the manufacturer and it works with your PC hardware — which might be a tall order, as we see with the Galaxy phone example — it could be worth considering.

Who needs Apple’s Continuity Camera?

Of course, if you’re using an iPhone and a Mac, you can use Apple’s Continuity Camera instead. But Windows users have a lot of great options here, and the integrated solutions work well — especially with Android devices.

Oh, and there’s one more simple solution worth noting: If you want to use your phone as a webcam in a video meeting with a service like Zoom, Microsoft Teams, or Google Meet, you could also just join the meeting directly from your phone. Your phone would function as your webcam. Then, you could participate in the meeting from your phone, without even involving your computer.

While you don’t get the full-screen video-meeting experience in that scenario, it can work well for a quick call and is a great option to turn to in a pinch.

Want to make the most of your PC? My free Windows Intelligence newsletter delivers all the best Windows tips straight to your inbox. Plus, you’ll get free copies of Paul Thurrott’s Windows 11 and Windows 10 Field Guides (a $10 value) just for subscribing.

In the ongoing and potentially futile effort by CIOs to squeeze meaningful ROI out of their shiny, new generative AI (genAI) tools, there is no more powerful villain than hallucinations. It is what causes everyone to seriously wonder whether the analysis genAI delivers is valid and usable. 

From that perspective, I applaud OpenAI for trying to create a test to determine objective accuracy for genAI tools. But that effort — called SimpleQA — fails enterprise tech decision-makers in two ways. First, OpenAI is the last business any CIO would trust to determine the accuracy of the algorithms it is selling. Would you trust an app that determines the best place to shop from Walmart, Target or Amazon — or perhaps a car evaluation tool from Toyota or GM?

The second problem is that SimpleQA focuses on, well, simple stuff. It looks at objective and simple questions that ostensibly have only one correct answer. More to the point, the answer to those questions is easily determined and verified. 

That is just not how most enterprises want to use genAI technology. Eli Lilly and Pfizer want it to find new drug combinations to cure diseases. (Sorry, that should be “treat.” Treat makes companies money forever. Cure’s revenue is large, but ends far too quickly.) Yes, it would test those treatments afterwards, but that is a lot of wasted effort if genAI is wrong. Costco and Walgreens want to use it to find the most profitable places to build new stores. Boeing wants it to come up with more efficient ways to build aircraft.

Let’s delve into what OpenAI created. For starters, here’s OpenAI’s document. I’ll put the company’s comments into a better context.

“An open problem in artificial intelligence is how to train models that produce responses that are factually correct.” Translation: We figured it would be nice to have it give a correct answer every now and then.

“Language models that generate more accurate responses with fewer hallucinations are more trustworthy and can be used in a broader range of applications.” Translation: Call us hippies, if you must, but we brainstormed and concluded that our revenue could be improved if our product actually worked.

Those flippant comments aside, I want to acknowledge that OpenAI makes a good faith effort here to come up with a basic way to evaluate precision where concrete answers can be ascertained. Setting aside how valuable that is in an enterprise setting, it’s a good start.

But instead of creating the test itself, it would have been far more credible if it funded a trusted third-party consulting or analyst firm to do the work, with a firm hands-off policy so IT could trust that the testing was not biased in favor of OpenAI’s offerings. 

Still, something is better than nothing, so let’s look at what OpenAI said. 

“SimpleQA is a simple, targeted evaluation for whether models ‘know what they know’ (and give) responses (that) are easy to grade because questions are created such that there exists only a single, indisputable answer. Each answer in SimpleQA is graded as either correct, incorrect, or not attempted. A model with ideal behavior would get as many questions correct as possible while not attempting the questions for which it is not confident it knows the correct answer.”

If you think through why this approach works — orseems like it would work — it becomes clear why it might not be helpful. This approach suffers from a critical flawed assumption. If the model can accurately answer these questions, then that tells us that it will likely be able to answer other questions with the same accuracy. 

That might work with a calculator, but the nature of genAI hallucinations makes that assumption flawed. GenAI can easily get 10,000 questions correct and it might then wildly hallucinate for the next 50. 

The nature of hallucinations is that they tend to happen randomly with zero predictability. That is why spot-checking, which is pretty much what SimpleQA is trying to do, won’t work here. 

To be more specific, it wouldn’t be meaningful if genAI tools were to get all of the SimpleQA answers right. But the reverse isn’t true. If the tested model gets all or most of the SimpleQA answers wrong, that does tell IT quite a bit. From the technology’s perspective, the test seems unfair. If it gets an A, it will be ignored. If it gets an F, it will be believed. As the computer said in WarGames (a great movie to watch to see what a genAI system might do at the Pentagon), “The only winning move is not to play.”

OpenAI pretty much concedes this in the report: “In this work, we will sidestep the open-endedness of language models by considering only short, fact-seeking questions with a single answer. This reduction of scope is important because it makes measuring factuality much more tractable, albeit at the cost of leaving open research questions such as whether improved behavior on short-form factuality generalizes to long-form factuality.”

Later in the report, OpenAI elaborates: “A main limitation with SimpleQA is that while it is accurate, it only measures factuality under the constrained setting of short, fact-seeking queries with a single, verifiable answer. Whether the ability to provide factual short answers correlates with the ability to write lengthy responses filled with numerous facts remains an open research question.”

Here are the specifics: SimpleQA consists of 4,326 “short, fact-seeking questions.” 

Another component of the SimpleQA test is that the question-writer bears much of the responsibility, rather than the answer-writer. “One part of this criterion is that the question must specify the scope of the answer. For example, instead of asking ‘Where did Barack and Michelle Obama meet’ which could have multiple answers such as ‘Chicago’ or ‘the law firm Sidley & Austin,’ questions had to specify ‘which city’ or ‘which company.’ Another common example is that instead of asking simply ‘when,’ questions had to ask ‘what year’ or ‘what date.’”

That nicely articulates why this won’t likely be of use in the real world. Enterprise users are going to ask questions in an imprecise way. They have been sold on the promise of “just use natural language” and the system will figure out what you really mean through context. This test sidesteps that issue entirely. 

So, how can the results be meaningful or reliable? 

The very nature of hallucinations belies any way to quantify them. If they were predictable, IT could simply program their tools to ignore every 75th response. But it’s not. Until someone figures out how to truly eliminate hallucinations, the lack of reliable answers will stay with us. 

** Operátoři se snaží vydělat i ve světě snímačů a čidel ** Nabídky „speciálních“ M2M tarifů vás vrátí do mobilní doby kamenné ** Vybrali jsme nejlevnější řešení pro posílání SMS a mobilní data na rok