What is generative AI?

Generative AI is a class of artificial intelligence utilities that leverage machine learning and neural networks to create completely new content. Today, generative AI can create anything from a painting of a rose in the style of Van Gogh, to a full screenplay. There are different types of generative AI, but for now, we’re going to focus on AI that generates text. 

The importance of generative AI

This class of artificial intelligence utilities is crucial in the next stages of human development. During the industrial revolution, humans were the workforce. In the next phase of human development, the world will, in all likelihood, be powered and kept running primarily by technology. The AI models that underpin generative AI have transformative capabilities in everything from research and word processing to complex sciences like computer engineering. 

The best AI text generators for 2023

Rytr

Rytr is an AI writing utility created to generate text and help you write better. Among other tasks, Rytr can create blog outlines and help you come up with ideas, as well as generate product descriptions and create social media ads. 

ContentBot

This is one of the preeminent AI text generators on the market. You can try ContentBot for free, but it is a premium tool. Billed as ‘the future of writing,’ this utility can create a whole host of different pieces of text, including SEO insights and reports, press releases, paraphrased content, and even various types of marketing copy. 

CopyAI

This is a free utility, and it’s built on one of the world’s most advanced AI language models. Founded in 2020, CopyAI can help you generate content with the help of 90+ unique tools. You can either create content with just one click, or you can simply use CopyAI’s tools to ensure that your writing is up to scratch and to the point. 

NeuroFlash 

This is another AI text generator that offers a free plan. NeuroFlash lets users write 2000 words for free each month. Included in the premium package are 90+ professional templates for various types of writing. The utility also has a tool that lets you write assorted blog posts in workflow with only 3 clicks. 

Simplified

With Simplified, you can write everything with AI. This is billed as the only marketing tool you’ll ever need, and it’s quite comprehensive. Suited particularly well to marketing, this utility lets you create compelling copy for ads, company descriptions, blog posts, eCommerce assets, and so much more. This is a feature-rich premium AI-powered text generator, and it makes coming up with content that much easier.

Source

Microsoft plans to lay off 10,000 workers, the company said Wednesday, as it looks to trim costs amid economic uncertainty and to refocus on strategic priorities, such as artificial intelligence.

The company employed about 221,000 workers as of the end of June, and the cuts amount to less than 5 percent of its global work force.

“These are the kinds of hard choices we have made throughout our 47-year history to remain a consequential company in this industry that is unforgiving to anyone who doesn’t adapt to platform shifts,” Satya Nadella, Microsoft’s chief executive, said in a message to staff.

The layoffs, which will begin on Wednesday, are the company’s largest in roughly eight years. Mr. Nadella cut about 25,000 jobs over the course of 2014 and 2015 as Microsoft abandoned its ill-fated acquisition of the mobile phone maker Nokia.

Like other tech companies, Microsoft expanded rapidly during the pandemic. It has hired more than 75,000 people since 2019, seizing on the surge in online services and the expansion of cloud computing.

Microsoft’s annual revenue grew 58 percent over three years, but rising interest rates and the prospect of a recession have tempered the company’s outlook. In the quarter that ended in October, it reported its slowest growth in five years and warned that more tepid results could follow.

The changes, including severance and other restructuring expenses, will cost $1.2 billion, Mr. Nadella said. Microsoft is scheduled to report its quarterly earnings on Tuesday.

The company has been pursuing several expensive bets, including potentially putting another $10 billion into its investment in OpenAI, which makes the explosively popular ChatGPT artificial intelligence system, and a $69 billion acquisition of the video game maker Activision that is facing challenges globally by antitrust regulators.

Other tech giants have also been reducing costs after several years of breakneck expansion. Amazon is expected to begin a huge round of layoffs on Wednesday as part of its plans to reduce its corporate work force by about 18,000 jobs.

The business software company Salesforce said this month that it planned to lay off 10 percent of its work force, or about 8,000 employees; and Meta, the parent company of Facebook, announced at the end of last year that it was cutting more than 11,000 jobs.

Source

The M2 Pro, which will be available in the just announced MacBook Pro and Mac mini, is built using 5-nanometer process technology, 20 percent more transistors than the M1 Pro, and double the amount in M2.

Key features of the M2 Pro include up to 200GB/s unified memory bandwidth, and up to 32GB of low latency unified memory. It also includes 10- or 12-core CPUs consisting of eight high-performance cores and four high-efficiency cores, granting a 20 percent performance increase over the M1 Pro.

The GPU in the M2 Pro can also be configured with up to 19 cores, three more than the M1 Pro, including a larger L2 cache and increasing graphics performance up to 30 percent over the previous generation.

The M2 Max on the other hand, is exclusive to the latest MacBook Pro models and is made with 67 billion transistors, 10 billion more than the M1 Max and 3x more than the M2. It comes with 400GB/s of unified memory bandwidth, twice that of the M2 Pro and 4x that of M2, and supporting up to 96GB of unified memory.

It includes the same 12-core CPU as the M2 Pro, but the main gains come to the GPU, with the M2 Max going up to 38 cores and is also paired with the larger L2 cache. This allows for graphics speeds to go up to 30 percent faster than M1 Max, which Apple claims to be the world's most powerful and efficient chip for a pro laptop.

Apple unveils next-generation M2 Pro and M2 Max chips, boosting CPU and GPU performance

Apple says the new processors deliver "pro-level performance" to the Mac mini, something "previously unimaginable" in such a compact design. Although the computer remains visually unchanged, it now offers faster processors, more unified memory, better connectivity, and up to three displays (on M2 Pro).

The Mac mini with 8-core M2 supports up to 24GB of unified memory with higher bandwidth, allowing faster performance in productivity and professional applications. According to Apple, image editing in Adobe Photoshop is up to 50% faster than in the previous generation. Also, the M2 processor enables ProRes acceleration for better video editing in Final Cut Pro with simultaneous playback of two 8K ProRes 422 30 fps or twelve 4K ProRes 422 video 30 fps streams.

Finally, no Apple presentation comes without throwing shade at Windows—Apple claims the new Mac mini is up to five times faster than the bestselling Windows desktop. Here are other performance comparisons Apple has provided:

When compared to the Mac mini with Intel Core i7,3 Mac mini with M2 offers:

• Up to 22x faster machine learning (ML) image upscaling performance in Pixelmator Pro.
• 9.8x faster complex timeline rendering in Final Cut Pro.

When compared to the previous-generation Mac mini with M1,1 Mac mini with M2 delivers:

• Up to 2.4x faster ProRes transcode in Final Cut Pro.
• Up to 50 percent faster filter and function performance in Adobe Photoshop.
• Up to 35 percent faster gameplay in Resident Evil Village.

The M2 Pro-based Mac mini has a 12-core CPU and 19-core GPU, plus up to 32GB of 200GB/s unified memory. There is also a "next-generation" Neural Engine with performance up to 40% faster than the M1. All this raw horsepower allows the computer to play up to five 8K ProRes 422 30 fps or 23 4K ProRes 422 video 30 fps streams.

When compared to the 27-inch iMac with Intel Core i7 and Radeon Pro 5500 XT,4 Mac mini with M2 Pro offers:

• Up to 50 percent faster filter and function performance in Adobe Photoshop.
• Up to 5.5x faster panoramic merge in Adobe Lightroom Classic.
• Up to 4.4x faster object tracking in Final Cut Pro.

When compared to the previous-generation M1 Mac mini,1 Mac mini with M2 Pro delivers:

• Up to 2.5x faster graphics performance in Affinity Photo.
• Up to 4.2x faster ProRes transcode in Final Cut Pro.
• Up to 2.8x faster gameplay in Resident Evil Village.

Connectivity-wise, the new Mac mini features four Thunderbolt 4 ports (two in the M2 variant), two USB-A, an HDMI, a Gigabit Ethernet port (with optional 10GB connectivity), an updated headphones jack that supports high-impedance headphones. Wireless includes Wi-Fi 6E and Bluetooth 5.3.

The updated Mac mini is now available for preorder on apple.com. Apple promises to ship the computer in select regions on January 24, 2023.

Apple's new Mac mini is now $100 cheaper, features M2 and M2 Pro chips

Today, Apple has announced updates to its 14-inch and 16-inch MacBook Pro models, originally launched in October 2021 with a complete redesign, more ports and the M1 Pro and M1 Max chips. These ugprades give it the latest M2 Pro and M2 Max chips, the second generation higher performance versions of their in-house silicon.

Apple touts that the newest chips are up to 6x faster than the highest specification Intel equivalent MacBook Pro, a battery life of up to 22 hours, advanced HDMI supporting up to two 8K displays at the same time, and the inclusion of Wi-Fi 6E support to give twice as fast connectivity.

Greg Joswiak, Apple's senior vice president of Worldwide Marketing, had the following to say:

“MacBook Pro with Apple silicon has been a game changer, empowering pros to push the limits of their workflows while on the go and do things they never thought possible on a laptop.”

“Today the MacBook Pro gets even better. With faster performance, enhanced connectivity, and the longest battery life ever in a Mac, along with the best display in a laptop, there’s simply nothing else like it.”

The key differences to the M2 Pro and M2 Max in comparison to the first-generation M1 Pro and M1 Max chips are primarily based around performance.

The M2 Pro contains a 10- or 12-core CPU, granting eight high-performance cores and four high-efficiency cores, which gives a performance increase of 20 percent over 2021's M1 Pro model. Apple also states that compiling in Xcode is up to 25 percent faster and image processing in Photoshop is up to 40 percent faster.

The M2 Max takes the graphics performance up to 38 cores, giving 30 percent greater graphics performance in comparison to the M1 Max, with up to 96GB of unified memory.

Despite the fact that Apple did not have a press event today similar to other product reveals, it has uploaded an 18 minute long keynote-style video to their YouTube channel with all of today's announcements.

Apple announces MacBook Pro range updates with M2 series of chips

A history of ARM, part 3: Coming full circle

For those unaware, tracking tags are items that you attach to devices or equipment such as keys, wallets, cars, or even your pets. This makes it easier to keep track of them and prevent theft. Some people have been misusing these items to stalk people, however, resulting in a lawsuit filed last December.

Wojciechowski recently discovered references showing that Google is working on support for locator tags in Fast Pair, a feature in Android that automatically discovers Bluetooth accessories in close proximity. They linked to a tweet by Mishaal Rahman of Esper:

According to Wojciechowski, the tracking tag apparently has the codename "grogu," and is also sometimes referred to as "groguaudio" or "GR10" within Google. The device, which will be developed by the Nest Team, will come with an onboard speaker for alerts, be available in a variety of colours, and offer support for ultra-wide band and Bluetooth Low Energy (LE). This helps ensure accurate item tracking without consuming a lot of battery power.

Wojciechowski currently doesn't know the exact date when Google will launch its own tracking tags. However, they did suggest that Google will probably announce the product at Google I/O Developer Conference and then launch it at the company's annual fall event, alongside new Google Pixel devices.

Source: Kuba Wojciechowski (Twitter)

Google is reportedly working on an Apple AirTags competitor

However, things could be set to change for the positive. At CES 2023, alongside many new processors and graphics cards for desktops and mobile, AMD also introduced new non-X Ryzen 7000 parts which are slightly slower variants than their X counterparts at a cheaper price point.

Alongside that, AMD is also looking to introduce the entry-level A620 chipset. New Gigabyte A620 filings have been spotted on the Eurasian Economic Commission (EEC)'s website today (via VideoCardz).

While EEC certification filings don't mean that these boards will be released immediately, it does show AMD's intent on bringing newer options to the low-budget segment of the market, and these should be out by the second half (2H) of 2023.

AMD Ryzen 7000 could finally become affordable as A620 AM5 makes first tiny mark

Images generated by artificial intelligence (AI) have come a long way in recent years. In 2014, for instance, a top-of-the-range industry-leading AI could just about produce an image that looked like it was taken on a Game Boy color. Now it's possible to type in a few words, and less than a few minutes later see Darth Maul getting hitched, or a kickass baby doing a solo skydive.

Star Wars characters getting married

Skydiving Baby

However, there are still a number of issues that keep cropping up – even if you ignore the demons Loab and Crungus that keep invading AI art uninvited. On Friday, one Midjourney enthusiast shared images that the AI had created of parties, none of which had taken place.

As well as a number of biases that Twitter user Miles encountered – including getting the AI to show "people who aren't white as mayonnaise" when asking it to depict "people" – Twitter users noticed some other details in the images which would make you believe in demons if the photos were verified to be real.

To cut a long story short, the AI produces nightmare fuel whenever it is asked to do teeth or hands.

At first glance, the photos may look quite like a party – but the moment you start counting fingers or teeth it all falls apart. By and large, people were not impressed by the overall effect.

"Thinking about going with an acquaintance to a cramped house party," illustrator Trevor Henderson wrote. "As the night goes on you notice people in the crowd grinning at you with mouths that hold too many teeth, hands grip your shoulders possessively, squirming with far too many fingers. You're the guest of honor."

AI, like anyone taking a high school art class, struggles with hands. 

"Due to the complex geometry of hands, there is no standard set of lines or shapes that AI can recognize as a hand," one designer and AI expert explained. "In order to generate realistic hands, AI needs to collate multiple different shapes and arrangements. There are about 30 points of geometric divergences in the human hand, from finger length and width to the metacarpals and wrist joints."

The AI works on pattern recognition and producing patterns, rather than possessing the concept of hands themselves. 

It may think it has nailed the pattern, the same with teeth, by putting an indeterminate amount of fingers clumped around a palm, but a quick glance by someone who has the concept of hands nailed (humans win again) and you are in uncanny valley territory.

Source

Lasers can act as virtual lightning rods, redirecting the direction in which bolts jump – although we’re a long way from being able to call down the wrath of science on unbelievers like a creature of myth.

The Franklin lightning rod was a major scientific advance of its day, preventing millions of fires and electrocutions and demonstrating humanity’s capacity to control forces we had long feared as belonging to the gods. Nevertheless, it’s been 270 years, and it remains the basis of our lightning protection: maybe it’s time for an upgrade.

That is what Dr Aurélien Houard of ENSTA Paris and co-authors propose in a new paper, demonstrating that laser pulses can change the direction of a lightning strike.

The team previously demonstrated lasers' capacity to ionize air in laboratories can cause 2-million-volt sparks to jump along low-density channels. To take their idea to a larger stage, they placed a laser machine the size of a car near a tower on Säntis Mountain, Switzerland. The tower was chosen as, contrary to sayings about lightning never striking twice in the same place, it gets hit about 100 times a year – reportedly the most in Europe.

During a six-hour storm, the laser pulses controlled the direction of four lightning discharges. One of the bolts, occurring under relatively clear skies, was recorded using two high-speed cameras. This showed it followed the laser beams’ path for at least 50 meters (164 feet). All were accompanied by increased X-Ray bursts.

“Although this research field has been very active for more than 20 years, this is the first field-result that experimentally demonstrates lightning guided by lasers,” the authors write. Two previous attempts to achieve something similar failed. The authors attribute their success to using much faster laser pulses, repeated on a scale of a thousand times a second.

While unquestionably cool, one may ask if this technique is all that practical. A laser like this will probably always be expensive to make and operate. Meanwhile, the telecommunications tower in question is still standing after all these years because its Franklin rod works extremely well on its own.

It’s certainly true that lightning guidance will not come cheap. It took the team three years to build their machine and its power was in the terawatt range, admittedly for a very short amount of time. This is more than the entire electricity consumption of Europe.

However, there are times when more than stationary rods are required, such as when people need to be moving around in an open field. Rockets with wire attached have been shown to trigger lightning, defusing strong electric fields, that could have led to subsequent strikes. However, the rockets are single-use and therefore possibly more expensive in the long run than lasers, which the authors argue could also protect vital stationary infrastructure better than Franklin rods.

Contrary to how we might imagine the process, the lightning all started at the top of the tower, with their upward path laser controlled, rather than a bolt from the heavens redirected to a different spot on the ground.

Also very useful if you want to power a time machine to get back to the future.

The paper is published in Nature Photonics. 

Source

Disclaimer: I have spent a considerable amount of time on these tools and others like them, and, unfortunately, there’s a massive issue with how they work. We’ll discuss the issue once I’ve reviewed the three best products for the job, but I just want you to enter this article with the knowledge that AI detection tools are not perfect. 

GPT-2 Output Detector

This first utility hails from the same stable as ChatGPT. Created by OpenAI, GPT-2 Output Detector is one of the foremost detection tools in the industry. OpenAI has been working on mimicking and duplicating human artistic ability and intelligence since 2015. The company went viral in 2022 on the back of the release of ChatGPT, a tool that inspired so many sleepless nights and cold sweats until I realized that it couldn’t replace me. 

However, while the company was working on the tool I thought would replace me, t was also working on a way to make sure that didn’t happen to anyone in my field. OpenAI created a way to make ChatPGT less usable in professional circles, and a way to detect if it had been used. 

You can use the online demo version of this tool right now, although we don’t know how long before it’s replaced by the final software. All you need to do is head to the website in the link above, paste your text into the box, and the tool will determine whether the text was written or generated. 

GLTR (Giant Language Model Test Room)

Another company that took rather diligent note of the OpenAI and its AI utilities is the MIT-IBM Watson AI Lab. In 2019, when OpenAI released GPT-2, the Lab joined forces with the Harvard Natural Language Processing Group to create a utility that could detect computer generated text. This tool is GLTR, and it’s relatively clever. See, it uses an approach known as ‘takes one to know one.’ This means that it uses its prowess as an AI to process a piece of text and determine each next word. If the word that GLTR predicts matches the actual text, it’s most likely generated by AI. If the text deviates and GLTR isn’t able to predict what comes next, the text is most probably written by a human. 

GPTZero

We’ve actually written about this one before. Edward Tian is a senior at Princeton University, and he has previously expressed great concern for students using tools like ChatGPT in their university assignments. Edward believes that professors would have a keen interest in a tool that makes it easier to determine whether any text has been written by human hands or generated by AI bots. Therefore, he created GPTZero.

Edward Tian is a very clever young individual, at only 22 years old. He explains that GPTZero works by detecting randomness in sentences. Two other words he used were ‘perplexity’ and ‘burstiness.’ I won’t pretend to understand how GPTZero works, but I can tell you it has issues, just like the rest of these tools, including those not on this list. 

The trouble with AI detectors

As alluded to previously, I have spent a lot of time with both ChatGPT and the detectors that are meant to make it more difficult to use. My method was to write a piece of text myself and generate one using ChatGPT. I then ran both pieces of text through as many AI detector models as I could find. The results were always the same. One after the other, the detector utilities identified the AI-generated text 100% of the time. This doesn’t sound like an issue, and it isn’t. The issue comes in with how my own work compared. 

Again, one after the other identified my work as either written entirely by AI or in large part. I became anxious and tried a few more pieces of my own original work, with largely the same results. 

The issue is that people like me write for algorithms. We write for humans first, otherwise, there’d be no joy in this job. But, we also write for algorithms. Algorithms determine whether our choices of words are too complex for the average audience, whether our sentences are too long, whether our articles are concise enough to keep the audience’s attention, and whether Google and other search engines will regard our articles favorably. We’re being taught, inadvertently, to write how algorithms and AI would prefer. And, when we use these detectors, it shows. The other aspect to keep in mind is that AI was trained in our writing. Therefore, if an AI’s generated text seems robotic and pleases the Google overlords, what does that say about those of us who inadvertently trained it? 

To summarize, AI detectors are really good at detecting AI. Unfortunately, they often detect AI in human hands.

Source

Since it's still in the Beta version, it's free for the first 25 images, and then you will have to pay to use it. All you need to do is enter one of the newbie's channels, type /imagine: prompt and type whatever comes to your mind. For example, if you type, let's say /imagine prompt: Ellaof Bookland, you will get four images looking like this:

It's visible that AI makes it because it looks like an artificial image. Still, it's fascinating how AI can generate amazing images just with a few words. However, there are some rules of conduct that you should obey. They don't tolerate any abusive content, even adult images or gore. Even though it can create anything that comes to your mind, be respectful to others. Also, be careful with AI tools since they can be used to create malware. Thus, always keep your data protected and secure.

Talented people work on the Midjourney, adding amazing stuff quite often. A huge collection of styles and prompts has been added to Midjourney. It's impressive because it can allow you to modify your own images, turning them into something more artistic, even if it's artificial. Guy called Javi Lopez shared a fantastic collection of all available prompts for Midjourney.

As you can see, all these prompts are available on Midjourney. They include prompts for different historical periods of a human lifetime and games, creatures, artists, Sci-Fi, Mediums, and many more. This has turned Midjourney into a desirable text-to-image AI. However, to use more of this amazon text-to-image AI, you need to pay a monthly subscription. As in other subscriptions, there are three types of subscriptions, offering services according to the price. The basic is 10$ monthly; the Standard is 30$ monthly, and the Corporate 600$ yearly.

I am not an artist by profession, but I have enjoyed playing with this text-to-image AI. Indeed, it can help artists get ideas, but let them be artists. This text-to-image AI has great potential to create amazing artificial images. This raises another question. Can this really be called art in a traditional way? This debate will be left for subsequent discussions.

Midjourney: a huge collection of styles and all the prompts

New File Explorer

We heard some pretty exciting news this week where we got our first look at what could be a new version of File Explorer for Windows 11. This "leak" came from a recent Dev Channel build 25272, indicating that a new home page is in the works with more details like Insights, Activities, Properties, Related Files, Conversations, Sharing status, and more. This is in addition to the recommended content that Microsoft is planning for File Explorer too. The good news is that the enthusiasts among us can enable the hidden UI through the process mentioned here, but know that the software is unstable in its current state.

Speaking of the Dev Channel, we did get build 25276 this week. It brings several new features like live kernel dumps in Task Manager, updates to cloud storage notifications - including new reminders for OneDrive -, updated dialog boxes, and more. Naturally, this was also followed by a Windows Server Preview build 25276 without a changelog.

A security feature that may intrigue some readers is that Microsoft will soon be disabling insecure SMB guest authentication fallbacks in Windows 11 Pro. In the same vein, the Redmond tech giant is also planning on killing off Microsoft Support Diagnostic Tool (MSDT) within the next couple years, likely due to the security threats it poses.

Some of you might also have noticed slow download and install speeds when updating to new Windows 11 Insider builds. And if that is the case, you will be relieved to know that Microsoft is actively investigating the problem.

The death of Windows 8.1

This week, Microsoft finally put Windows 7 and 8.1 to sleep forever in terms of support. Both the operating systems are no longer supported which means that Microsoft will not provide technical support for them and you might run into security and usability issues while running the OSes too. It is important to note that Windows 7 was supported up until this week only for paid Extended Security Updates (ESUs) customers anyway.

Naturally, this was followed by news of several third-party software dropping support for the operating systems too. Among these are Paint.NET 5 and Google Chrome. That said, the latest version of Chrome is the last version to support the legacy OSes, support will officially end next month with Chrome 110. Amid all of this commotion, we learned that Microsoft mysteriously implemented a half-baked version of native UEFI and Secure Boot in Windows 7, which is odd considering the OS was pretty much on its deathbed.

Both Windows 7 and 8.1 also received their last Patch Tuesday updates this week. They pack some authentication improvements, fixes for a buggy SQL Server driver, and a couple of known issues. Meanwhile, Windows 10 nabbed some security fixes along with a known issue. And Windows 11 had similar improvements with two known problems. Interestingly, the Windows 11 Patch Tuesday also packs two hidden features (find out how to enable them here), but does not document them.

In terms of other fixes and improvements from Microsoft, the company resolved an OOBE issue in Windows 11 version 22H2 and a BSOD problem on Windows 10. It also established a partial fix for a Friday the 13th bug that led to Defender randomly deleting shortcuts on various versions of Windows. In the same vein, Microsoft is making it easier for IT admins to manage Win32 app upgrades and replacements across their organizations too.

Third-party developers weren't far behind in providing software updates either. Rufus received an updated FIDO script to fix broken Windows ISO downloads while its alternative Ventoy resolved numerous bugs related to Windows 11 requirements bypass, a VHDX booting issue, and more.

Finally, in terms of app updates, it seems like the new Windows 11 Media Player is also on its way to Windows 10. But if you simply can't wait for that, another interesting news item for you could be that hell has frozen over as preview apps for Apple Music, Apple TV, and Apple Devices have landed in the Microsoft Store.

Microsoft Teams paywall

Earlier in the week, Microsoft Teams customers received some not-so-good news as we reported that Microsoft is locking some existing features behind the more expensive Teams Premium add-on, launching next month at $10/user/month. The capabilities that will find themselves behind a paywall next month include live translated captions, custom Together mode scenes for organizations, timeline markers in Teams meetings recording marking when a member joined or left a meeting and more.

OneNote customers will be happy though with AI-backed voice commands for OneNote on Mac, along with preview features consisting of a keyboard shortcut for drawing straight lines, complete with Pen & Ink improvements. Similarly, Microsoft Planner finally netted Recurring Tasks and Grid view, while SwiftKey for Android offered a new way to change your app language. Microsoft also confirmed a bunch of issues in Outlook which it is still investigating. That said, it has offered a workaround for now.

This week also saw the release of Microsoft Edge 109 with security fixes, text prediction, TLS server certificate verification changes, and account Linking between a personal Microsoft account (MSA) and Azure Active Directory (AAD) account. In addition, we learned that Microsoft is working on even more rounded corners, this time in Edge.

Lastly, if you think that Microsoft 365 is too expensive for you, Microsoft is now offering an enticing Basic plan with the following capabilities:

• 100GB of space in OneDrive
• Ad-free Outlook on the web and mobile
• Additional security features, such as data encryption in Outlook, attachment scanning, and suspicious links checking
• Additional features for OneDrive, such as Personal Vault, password-protected and expiring links, ransomware data recovery, and bulk file store (coming later this year)
• Additional technical support over phone or online chat for apps and Windows 11

Microsoft 365 Basic costs $1.99/month or $19.99/year.

Git gud

Microsoft finally confirmed its rumored gaming showcase with Bethesda a couple of days ago. It's called "Developer_Direct" and is expected to offer closer looks at Forza Motorsport, Minecraft Legends, Redfall, and the future of The Elder Scrolls Online on January 25. Starfield is a noticeable absentee but Microsoft has assured gamers that a standalone showcase for the highly anticipated title is planned for later.

In other interesting news, Microsoft's Activision Blizzard acquisition deal is now facing opposition from Google and Nvidia too. The two have lodged complaints with the Federal Trade Commission (FTC) citing unfair advantages for Microsoft in the cloud, subscription, and mobile gaming sectors if the deal were to be approved.

This has not distracted Microsoft from other gaming areas. It recently published a video emphasizing the benefits of gaming on Windows 11 version 22H2, and made sure that Xbox is the first carbon aware console in the world.

On the topic of games, I Am Fish is the title on offer via the Xbox Free Play Days promotion while Mortal Shell: Enhanced Edition, Monster Hunter Rise, and a couple of Persona titles are the latest additions to Game Pass. But if none of that interests, you might want to check out this Weekend's PC Game Deals instead, curated by our News Editor Pulasthi Ariyasinghe.

Dev Channel

• Surface Laptop Studio now has better Windows Hello facial recognition
• Surface Pro 8 has netted a bunch of new drivers in the latest firmware update
• Microsoft has acquired Fungible in push for datacenter innovation
• Microsoft could be planning to invest $10 billion in OpenAI, with some caveats
• Android 13-based Windows Subsystem for Android is now available for Windows Insiders
• Latest Exchange Server updates have improved the security of PowerShell payloads but introduced a known issue too

Under the spotlight

News Reporter Taras Buria published a listicle with his thoughts on the top five features Microsoft added to Windows 11 in 2022. Find out whether or not you agree with him here.

Meanwhile, News Reporter Rahul Naskar shared his thoughts on how despite switching to macOS, he still misses Windows a lot.

On the other hand, News Reporter Hemant Saxena published a piece listing five ways to speed up your Windows PC if it has recently been struggling in terms of performance.

Lastly, if you're struggling to keep track of your tasks and to-do items, make sure to check out forum member Adam Bottjen's Tech Tip Tuesday piece about note-taking on Windows PCs.

Logging off

We actually have more than one interesting news items of the week in this section this time. For starters, the U.S. Congress has denied the Army's request for a $400 million fund to purchase almost 7,000 HoloLens headsets. The rationale behind this is that the investment is not worthwhile in the current state of the headsets with soldiers experiencing eye strain, nausea, and headaches. Instead, a $40 million budget has been approved to improve the headset to resolve the existing issues in the next iteration of the hardware.

Microsoft's hardware woes just don't end there though. The company has apparently ditched the Surface Duo 3 initially planned for late 2023, at least in its current form factor. This is due to the lukewarm reception to this lineup. Instead the firm is now working on a conventional foldable and another one with a "standard" smartphone design. This is part of the "Perfect Together" strategy to differentiate Surface hardware from other competitors. Only time will tell how successful this plan ends up being.

Microsoft Weekly: New File Explorer, the death of Windows 8.1, and a Teams paywall

The list below is all thanks to @thesyedhuq on Twitter, who recently went on a ChatGPT deep dive, exploring every facet of its capabilities. With the appropriate credit given, let’s get into it, shall we?

Get prepped for job interviews

Did you know that you can ask ChatGPT what the most common job interview questions are? It’s probably best to include your prospective industry in the prompt, but this utility is a brilliant way to familiarize yourself with the kind of questions that employers in a particular industry are likely to ask. This is also, as Syed notes, a brilliant way to gain a little confidence before the big day and prime your knowledge. 

Personal tutoring, anyone?

If you’re itching to get more acquainted with a particular topic, field of inquiry or industry, but you need a tutor who isn’t going to lose patience with you, ChatGPT is the best place to start. ChatGPT is trained on innumerable quantities of data; it can therefore act as a tutor in largely any subject. Whether you need help on a math quiz or someone to help you strengthen an essay argument, try ChatGPT. 

Write a speech

I’m a writer, and I’ve tested various AI writing assistants, so I’m a little dubious about this one. However, Syed asserts that ‘ChatGPT can help you brainstorm a speech or fill the gaps on an existing one to make it more impactful. It's like having a virtual assistant bounce ideas with you for a rough draft.’ 

I’m in agreement about the rough draft idea. This software is absolutely not good enough to write your essays for you. You’ll have to edit the software out, and yourself in, and that’s quite a big job, depending on the quantity of your writing. But it’s easy to use ChatGPT for ideas and inspiration. 

Brainstorm a side hustle

It’s increasingly difficult to make sufficient money from just a 9-5, unless you’re highly skilled and in demand. Therefore, it’s more important than ever for people to find side hustles with which to generate extra income. ChatGPT is really good, as it turns out, at giving you decent ideas about the viability of various side hustles. You can even ask the utility about the pros and cons of each. 

Summarize content

There simply isn’t enough time to read every exciting book, thesis, research paper, or abstract. And, unfortunately, when you have a sponge for a brain like mine, there are just so many of these works that you find interesting. Lighten your load a little by asking ChatGPT to give you a summary of a particular written text so that you can get the gist without reading the entire work. 

Generate SQL queries

I’ll let Syed take the lead on this one. ‘Data analysts rejoice, your lives are about to get easier. ChatGPT can help you write SQL queries from a simple text prompt. While you still need to verify the output, you can improve your workflow if you're just starting out.’

Fix coding errors

Coding is hard. Not speaking from personal experience here, but I know a few hardcore code-monkeys as The Sims 4 puts it, and I’ve heard them say ‘coding is hard.’ My sims complain from time to time to but they’re easy to persuade otherwise. If you’re struggling with a piece of code, you can run it through ChatGPT and ask it to offer up fixes. ChatGPT is also good for writing basic code, but it's nowhere near to replacing devs, so don’t stress. 

Personalized advice

Everyone needs advice sometimes. For the times that you have a problem but don’t necessarily want to consult friends or family about it, try ChatGPT. Think of this as a disinterested third party that is basically obligated to listen to your sadness, confusion, or heartbreak because that’s what it was coded to do. How perfect. 

Perform calculations

If you’re anything like me, you’re probably really bad at math. Unless you’re a coder. On the off-chance that you are as atrociously bad at calculations as I am, this will benefit you. ChatGPT was designed to be good at many things, including math. So if you need to solve for X or any other math-related stuff I can’t remember, use ChatGPT. 

Learn a new language

This is right up my alley, as I’m currently learning Greek, Norwegian, and German. It’s not my fault. My family keeps moving to Europe, and I’m planning ahead for the next one. ChatGPT is excellent for learning spelling, punctuation, and grammar in a new language. The tool doesn’t speak, so you won’t learn pronunciation. But it’s fully multilingual and can translate largely any language into largely any other language. 

Disclaimer

If you use ChatGPT for research, please fact-check the ‘facts’ it spits out. Software makes mistakes. Like when it claimed that elephants lay the biggest eggs in the animal kingdom. Additionally, you shouldn’t be using ChatGPT for professional tasks, especially not final products. The software just isn’t good enough yet. Still, it’s a really fun tool to play with, so have at it and have fun.

Source

The famous chatbot is able to create text responses to human prompts, even if those prompts ask it to write code. This means that among all of its more traditional literary skills, ChatGPT is also somewhat of a convincing coder. Convincing is the right word here, however, as the large language model has a problem with accuracy. Based on what it can learn from the vast amounts of data used to train it, it can create convincing texts and convincing code, that can look the part. However, it has no actual conception of what is right or wrong, either factually in the text or that the code is correct, which means users can’t know either.

Now there are plenty of users who won’t bother checking whatever ChatGPT spits out at them but unfortunately, we can’t rely on cybercriminals being quite so lazy. According to a new report by Check Point Research, they have spotted several posts on hacking forums, where malicious actors discuss all manner of nefarious activities, discussing ChatGPT and how best to use it. One of the more shocking revelations is that many of the users on the forums who are discussing ChatGPT in this manner are doing so from new accounts, which could indicate that the tool is allowing new users to get into cybercrime, even if previously they haven’t had the ability to do so. We discussed this very possibility a month ago and hope it didn’t give any of you any ideas ?.

In all seriousness, however, this does mark a dangerous escalation and could result in a flood of new cybercrime attacks spreading across the internet. The Check Point report does say that the malicious code it has been discovering mostly seems to be coming from rookies, but it clearly highlights the risk of more sophisticated actors bringing it into their workflows to augment the work they are already doing. The report says:

“t’s still too early to decide whether or not ChatGPT capabilities will become the new favorite tool for participants in the Dark Web. However, the cybercriminal community has already shown significant interest and is jumping into this latest trend to generate malicious code.”

If this trend does pick up, there may be hope in the form ChatGPT watermark, which is currently being developed to embed texts using cryptography with machine-readable watermarks. However, in its current form that still has shortcomings, and it has only ever been discussed relating to text output and not code.

Source

As a neuroscientist specializing in the brain mechanisms of consciousness, I find talking to chatbots an unsettling experience. Are they conscious? Probably not. But given the rate of technological improvement, will they be in the next couple of years? And how would we even know?

Figuring out whether a machine has or understands humanlike consciousness is more than just a science-fiction hypothetical. Artificial intelligence is growing so powerful, so quickly, that it could soon pose a danger to human beings. We’re building machines that are smarter than us and giving them control over our world. How can we build AI so that it’s aligned with human needs, not in conflict with us?

As counterintuitive as it may sound, creating a benign AI may require making it more conscious, not less. One of the most common misunderstandings about AI is the notion that if it’s intelligent then it must be conscious, and if it is conscious then it will be autonomous, capable of taking over the world. But as we learn more about consciousness, those ideas do not appear to be correct. An autonomous system that makes complex decisions doesn’t require consciousness.

What’s most important about consciousness is that, for human beings, it’s not just about the self. We see it in ourselves, but we also perceive it or project it into the world around us. Consciousness is part of the tool kit that evolution gave us to make us an empathetic, prosocial species. Without it, we would necessarily be sociopaths, because we’d lack the tools for prosocial behavior. And without a concept of what consciousness is or an understanding that other beings have it, machines are sociopaths.

Computer science pioneer Alan Turing in 1951.Photo: Godfrey Argent Studio/The Royal Society

The only diagnostic tool for machine consciousness that we have right now is the Turing test, a thought experiment named for the British computer scientist Alan Turing. In its most common version, the test says that if a person holds a conversation with a machine and mistakes its responses for those of a real human being, then the machine must be considered effectively conscious.

The Turing test is an admission that the consciousness of another being is something we can only judge from the outside, based on the way he, she or it communicates. But the limits of the test are painfully obvious. After all, a pet dog can’t carry on a conversation and pass as a human—does that mean it’s not conscious? If you really wanted a machine to pass the test, you could have it say a few words to a small child. It might even fool some adults, too.

The truth is, the Turing test doesn’t reveal much about what’s going on inside a machine or a computer program like ChatGPT. Instead, what it really tests is the social cognition of the human participant. We evolved as social animals, and our brains instinctively project consciousness, agency, intention and emotion onto the objects around us. We’re primed to see a world suffused with minds. Ancient animistic beliefs held that every river and tree had a spirit in it. For a similar reason, people are prone to see faces in random objects like the moon and moldy toast.

The original test proposed by Alan Turing in a 1950 paper was more complicated than the version people talk about today. Notably, Turing didn’t say a word about consciousness; he never delved into whether the machine had a subjective experience. He asked only whether it could think like a person.

Turing imagined an “imitation game” in which the player must determine the sex of two people, A and B. One is a man and one is a woman, but the player can’t see them and can learn about them only by exchanging typed questions and answers. A responds to the questions deceitfully, and wins the game if the player misidentifies their sex, while B answers truthfully and wins if the player identifies their sex correctly. Turing’s idea was that if A or B is replaced by a machine, and the machine can win the game as often as a real person, then it must have mastered the subtleties of human thinking—of argument, manipulation and guessing what other people are thinking.

Turing’s test was so complicated that people who popularized his work soon streamlined it into a single machine conversing with a single person. But the whole point of the original test was its bizarre complexity. Social cognition is difficult and requires a theory of mind—that is, a knowledge that other people have minds and an ability to guess what might be in them.

If we want to know whether a computer is conscious, then, we need to test whether the computer understands how conscious minds interact. In other words, we need a reverse Turing test: Let’s see if the computer can tell whether it’s talking to a human or another computer. If it can tell the difference, then maybe it knows what consciousness is. ChatGPT definitely can’t pass that test yet: It doesn’t know whether it’s responding to a living person with a mind or a disjointed list of prefab questions.

A sociopathic machine that can make consequential decisions would be powerfully dangerous. For now, chatbots are still limited in their abilities; they’re essentially toys. But if we don’t think more deeply about machine consciousness, in a year or five years we may face a crisis. If computers are going to outthink us anyway, giving them more humanlike social cognition might be our best hope of aligning them with human values.

Dr. Graziano is a professor of psychology and neuroscience at Princeton University and the author of “Rethinking Consciousness: A Scientific Theory of Subjective Experience.”

Source

“The Compensation Committee balanced shareholder feedback, Apple’s exceptional performance, and a recommendation from Mr Cook to adjust his compensation in light of the feedback received,” Apple wrote in the filing. “The Compensation Committee then approved the target annual compensation of our named executive officers for 2023, including the following changes to CEO compensation, which Mr Cook supported, and that the Compensation Committee believes are responsive to shareholder feedback, while continuing both to align pay with performance and to recognize Mr Cook’s outstanding leadership.”

While Cook’s target compensation is falling from $84 million to $49 million, there will be no change to his $3 million base pay, nor his $6 million annual cash incentive. Instead, his equity award value will fall from $75 million to $40 million and the award mix (performance-based, time-based) will move from 50/50 to 75/25.

Despite the $84 million target compensation mentioned earlier, the total compensation paid out to the Apple CEO totalled $99.4 million in 2022, when factoring in things like air travel and security. In the same year, the median compensation for Apple employees was $84,493 so Cook’s pay ratio was 1,177 to 1, according to the document.

While some may argue that this is an obscene amount of money to be earning each year, it’s worth noting how much Apple's worth has increased since Cook took over from Steve Jobs. In the last five years alone the price of Apple’s shares has increased by 198.99% and that’s despite a pullback over the course of 2022. When Cook took over, Apple’s shares were in the low double digits but now they sit at $133. He's likely to be doing something right at the head of the company and arguably deserves the pay he gets.

Source: SEC

Apple CEO Tim Cook's pay slashed by 40% for 2023

This information was divulged to Bloomberg by sources it has kept anonymous. They say that Google and Nvidia have told the FTC that Microsoft could gain an unfair advantage from the deal in the cloud, subscription, and mobile gaming sectors if the deal were to be approved. The sources revealed that Nvidia isn’t strictly against the acquisition but wants to ensure there is equal and open access to game titles like Call of Duty.

The FTC is due to hold an in-house trial in August regarding the acquisition. Google, Nvidia, or Sony may be called to testify in the FTC’s trial. Any compromises that allow the deal to go ahead will likely include forcing Microsoft to keep access to the games open for at least a decade. Microsoft has already expressed a willingness to do this but it’s unclear what would happen after that.

Call of Duty is the game most of Microsoft’s rivals will be most concerned about due to its popularity but Activision Blizzard owns lots of other really popular titles too such as World of Warcraft, Crash Bandicoot, Guitar Hero, StarCraft, Hearthstone, Diablo, Overwatch, Candy Crush Saga, and more. Hopefully, regulators can find a way to ensure the acquisition works for all the concerned parties.

Source: Bloomberg

Google, Nvidia raise concerns over Microsoft's $69 billion Activision Blizzard acquisition

Why do old photos look blurry?

For starters, you’re not losing your mind. While that photograph of you from the 80s itself hasn’t changed in quality, our perception of such media has. When you looked at the photo in the 80s, it was a pretty rad pic. You were proud. Everyone was recognizable, and none of you could believe how clear and crisp it was. The technology back then was amazing - but only back then. Since then, technology has improved 100-fold. 

Related: AI is better at determining gestational age.

We have a far higher expectation of photos today. If you can’t see someone’s pores and their fear of their future in their eyes, it’s not a good enough photo. There was nothing wrong with your predecessor's skills with a camera, the camera was simply just not good enough for what we deem acceptable today. Luckily, there’s a tool to turn your old photographs into ones you can haul out proudly at family dinners, without anyone squinting awkwardly or asking ‘Who’s that?’ 

The tool that we are referring to is called RestorePhotos.io and it is truly incredible. Using the advanced capabilities of AI image recognition and regeneration software, the utility scans your old blurry photographs and gives them a crisp edge. But, that’s just my inference. Here’s what the company behind it, Vercel, says about it. 

‘It uses an ML model from the Applied Research Center called GFPGAN on Replicate to restore face photos. This application gives you the ability to upload any photo, which will send it through this ML Model using a Next.js API route, and return your restored photo. After cloning the repo, go to Replicate to make an account and put your API key in a file called .env. If you'd also like to do rate limiting, create an account on UpStash, create a Redis database, and populate the two environment variables in .env as well. If you don't want to do rate limiting, you don't need to make any changes.’ 

Related: Adobe is training its AI on user Data

If you’re not the most tech-savvy person and the paragraph above flew over your head, don’t worry. All you really need to know is that using RestorePhotos.io is completely free, and that it can bring life back into old photographs and keep those memories fresh.

Source

Intel breaks the 6GHz barrier with $699 Core i9-13900KS processor

The Samsung PM9C1a SSD aims to improve two primary aspects of a PC drive: speed and efficiency. The new controller delivers 60% faster sequential read speed and 80% sequential write speed, allowing the drive to operate at 6,000MB/s and 5600MB/s, respectively. Random input-output operations also received a significant boost: Samsung claims the drive can run at 900K IOPS read and 1,000K IOPS write. These speeds are not record-breaking by any means (Samsung's 990 PRO delivers faster performance), but they are plenty enough for daily use at a much lower price.

As for energy efficiency, the PM0C1a features up to 70% better efficiency per watt than its predecessor (PM9B1). This allows laptops and tablets with the new SSD to consume less power while handling the same amount of tasks without performance trade-offs. Besides, the drive now requires approximately 10% less energy when a PC goes into standby mode.

Customers can expect the Samsung PM0C1a to appear inside new laptops and tablets with 256GB, 512GB, and 1TB capacities. Form factors include M2 2230, 2242, and 2280.

Samsung unveils next-gen PCIe 4 SSD with new 5nm controller and speeds up to 6000MB/s

Academics can be fooled into believing bogus scientific abstracts generated by ChatGPT are from real medical papers published in top research journals, according to the latest research.

A team of researchers led by Northwestern University used the text-generation tool, developed by OpenAI, to produce 50 abstracts based on the title of a real scientific paper in the style of five different medical journals.

Four academics were enlisted to take part in a test, and were split into two groups of two. An electronic coin flip was used to decide whether a real or fake AI-generated abstract was given to one reviewer in each group. If one researcher was given a real abstract, the second would be given a fake one, and vice versa. Each person reviewed 25 scientific abstracts.

Reviewers were able to detect 68 per cent of fake abstracts generated by AI and 86 per cent of original abstracts from real papers. In other words, they were successfully tricked into thinking 32 per cent of the AI-written abstracts were real, and 14 per cent of the real abstracts were fake.

Catherine Gao, first author of the study and a physician and scientist specialising in pulmonology at Northwestern University, said it shows ChatGPT can be pretty convincing. "Our reviewers knew that some of the abstracts they were being given were fake, so they were very suspicious," she said in a statement.

"The fact that our reviewers still missed the AI-generated ones 32 [per cent] of the time means these abstracts are really good. I suspect that if someone just came across one of these generated abstracts, they wouldn't necessarily be able to identify it as being written by AI."

Large language models like ChatGPT are trained on large amounts of text scraped from the internet. They learn to generate text by predicting what words are more likely to occur in a given sentence, and can write grammatically accurate syntax. It isn't surprising that even academics can be fooled into believing AI-generated abstracts are real. Large language models are good at producing text with clear structure and patterns. Scientific abstracts often follow similar formats, and can be quite vague.

"Our reviewers commented that it was surprisingly difficult to differentiate between the real and fake abstracts," Gao said. "The ChatGPT-generated abstracts were very convincing…it even knows how large the patient cohort should be when it invents numbers." A fake abstract about hypertension, for example, described a study with tens of thousands of participants, whilst one on monkeypox included a smaller number of patients.

Gao believes tools like ChatGPT will make it easier for paper mills, who profit from publishing studies, to churn out fake scientific papers. "If other people try to build their science off these incorrect studies, that can be really dangerous," she added.

There are advantages to using these tools too, however. Alexander Pearson, co-author of the study and an associate professor of medicine at the University of Chicago, said they could help non-native English scientists write better and share their work.

AI is better at detecting machine text than humans. The free GPT-2 Output Detector, for example, was able to guess with over 50 per cent confidence that 33 out of 50 papers were indeed generated by a language model. The researchers believe paper submissions should be run through these detectors, and that scientists should be transparent about using these tools.

"We did not use ChatGPT in the writing of our own abstract or manuscript, since the boundaries of whether this is considered acceptable by the academic community are still unclear. For example, the International Conference on Machine Learning has instituted a policy prohibiting its use, though they acknowledge that the discussion continues to evolve and also clarified that it is okay for it to be used in 'editing or polishing'," Gao told The Register.

"There have been groups who have started using it to help writing, though, and some have included it as a listed co-author. I think that it may be okay to use ChatGPT for writing help, but when this is done, it is important to include a clear disclosure that ChatGPT helped write sections of a manuscript. Depending on what the scientific community consensus ends up being, we may or may not use LLMs to help write papers in the future." ®

Source

Although TSMC's N3 (3nm-class) family of fabrication processes brings a number of benefits in terms of performance and power, the very high costs of the foundry's initial N3 node hampers widespread adoption. Unsurprisingly, the company is rumored to be preparing to lower its quotes for 3nm production to stimulate interest from chip designers, according to a report from MyDrivers. 

While at this point any published TSMC's N3 quotes and prices should be considered rumors, it's likely that TSMC's production costs on its N3E process will be lower than those on its initial N3. It remains to be seen how much the company will charge for production on other N3-class nodes, such as N3P, N3S, and N3X. Lowering prices of 3nm production will attract more customers to these nodes, but this is not something that is going to happen overnight. 

TSMC's initial N3 manufacturing technology (also known as N3B) is rumored to be used only by Apple because the company is the foundry's largest client willing to adopt leading-edge nodes ahead of others. But N3 is an expensive technology to use. N3 extensively uses extreme ultraviolet (EUV) lithography for up to 25 layers, according to China Renaissance, and each EUV scanner now costs $150 million - $200 million, depending on configuration. To depreciate fabs equipped with such production tools, TSMC has to charge more for production on its N3 process and successors.  

Some say that TSMC might be charging as much as $20,000 per N3 wafer — up from $16,000 per N5 wafer — and while such quotes depend on numerous factors, the key takeaway is that chip production keeps getting more expensive. Increased costs mean lower profits for companies such as AMD, Broadcom, MediaTek, Nvidia, and Qualcomm, which is why chip developers are reconsidering how they create advanced designs and use leading-edge nodes. 

"We believe the meaningful [N3] ramp-up will be in 2H 2023 when the optimized version, N3E, will be ready," wrote Szeho Ng, an analyst with China Renaissance. "Its major customers in HPC (i.e., AMD, Intel), smartphone (i.e., QCOM, MTK) and ASIC (i.e., MRVL, AVGO, GUC) will likely stay in N4/5 and choose N3E as their maiden N3 class foray, in our view. Meanwhile, we believe the baseline N3 (aka N3B) adoption will be largely limited to Apple products." 

To stimulate its partners into using its N3-class process technologies, TSMC is reportedly considering lowering its quotes for these nodes. In particular, TSMC's N3E process uses EUV only for up to 19 layers and features somewhat lower complexity in terms of manufacturing, and is thus cheaper to use. TSMC could lower quotes of N3E production without harming profitability. N3E provides zero advantages over N5 when it comes to SRAM cell scaling, which means larger die sizes when compared to those made on N3/N3B. 

AMD publicly announced that it planned to use an N3 node for some of its Zen 5-based designs due in 2024, and Nvidia is expected to adopt N3 for its next-generation Blackwell architecture-based GPUs set to arrive around the same timeframe. Due to high costs, adoption of N3-class nodes is expected to be limited to certain products — so lowering quotes will probably make chip designers reconsider their adoption strategy.  

There's also another issue with TSMC's N3: low yields. Some estimate yields are between 60% and 80%, and sources at DigiTimes (via Dan Nystedt) indicate that they're below 50%. That said, since only Apple reportedly uses this manufacturing technology and the company is known for being very secretive, any details about yields of initial N3 chips should be taken with a large grain of salt.

Source

The experimental ideas might sound like sci-fi, but they just received a bundle of funding from NASA.

Planes built for Saturn's largest moon, self-growing building blocks, deep space satellite mega-constellations, and oxygen pipelines on the Moon are just a few of the 14 experimental projects selected by NASA to receive more of their funding in a push to advance the agency’s exploration of space. As part of the Innovative Advanced Concepts (NIAC) program, each of these bold research projects will receive $175,000 in grants to complete a Phase 1 study of their idea. 

NASA says the projects are still in the “very early stages of conceptual development,” but they hope the small boost of money could see which of these high-risk, high-reward ideas has the potential to get off the ground. 

“NASA dares to make the impossible possible. That’s only achievable because of the innovators, thinkers, and doers who are helping us imagine and prepare for the future of space exploration. The NIAC program helps give these forward-thinking scientists and engineers the tools and support they need to spur technology that will enable future NASA missions,” Bill Nelson, NASA Administrator, said in a statement. 

One of the selected initiatives is the Washington-based Planet Enterprises for their vision of creating a plane capable of complete missions to Saturn's moon Titan. Titan is a fascinating icy world with a dense golden atmosphere that has the potential to harbor extraterrestrial life,

The “TitanAir” mission would involve a seaplane-like vehicle that’s capable of gliding around the moon and picking up condensation of methane and suspended complex organics. It could also land on the icy lakes for the purposes of scooping up liquid samples. 

Graphic depiction of the Lunar South Pole Oxygen Pipeline. Image credit: Peter Curreri/NASA

Another project that sounds straight out of science fiction comes from Lunar Resources, Inc in Houston. Their idea is to install an oxygen pipeline on the surface of the Moon to serve future human settlements there. 

They hope to extract oxygen from dust and water from lunar ice at the moon’s South Pole and then use pumps down a 5-kilometer-long (over 3-mile) pipeline towards human habitats, rovers, and launch vehicles that would require it. 

Researchers at Massachusetts Institute of Technology put forward the plan of the Great Observatory for Long Wavelengths (GO-LoW) that hopes to deploy thousands of satellites into space at points between Earth and the Sun. Their job will be to measure the magnetic fields of terrestrial exoplanets via detections of their radio emissions.

The University of Nebraska earned its NASA grant by proposing self-growing building blocks that could be used to build habitats on Mars. Their proposal is to use cyanobacteria and fungi as building agents that will pump out biominerals, like calcium carbonate, and biopolymers that could be used as sturdy building materials. 

ZZZZZZAP. Artist's depiction of the EmberCore Flashlight. Image credit: Christopher Morrison/NASA

The EmberCore Flashlight was put forward by the Ultra Safe Nuclear Corporation. Their plan is essential to develop a high-tech “flashlight” that can illuminate dark regions of the moon using X-rays/gamma-rays. This, they hope, could help to reveal the composition of the geology in the bedrock of the "dark side" of the moon.

Source

IDC compares the shipments in the fourth quarter of 2022 to the same period in 2018 where the market was constrained by Intel's supply challenges. The company also said that while it was clear that the PC market's pandemic buying boom is over, PC shipments for 2022 were well above pre-pandemic levels at 292.3 million units.

"Consecutive quarters of declines clearly paint a gloomy picture of the PC market, but this is really all about perception," said Ryan Reith, group vice president with IDC's Worldwide Mobility and Consumer Device Trackers. "2021 was near historic levels for PC shipments, so any comparison is going to be distorted."

Despite the decline, IDC believes that the PC market has the potential to bounce back in 2024 and the company sees "pockets of opportunity" throughout 2023.

When it came to manufacturers, Lenovo is still the top PC maker, shipping 15.5 million units in the fourth quarter of 2022. HP came in second with 13.2 million units sold, a drop of about 5.4 million units from its fourth quarter sales last year. Dell completes the trio, selling 6.4 million fewer units than last year. While Apple's PC shipments also declined, they are the only one to record a single digit drop at -2.1%.

IDC: PC shipments dropped by 28% in the 4th quarter of 2022