So while AM4 and compatible previous-gen Ryzen chips continue to be great inexpensive options, AMD has been slowly dropping Ryzen 7000 prices too. And now, affordable AM5 motherboards have begun arriving as well as AMD's David McAfee, the CVP and GM of Ryzen channel, pointed it out today with Newegg offering the first $125 B650 motherboard.

The board is a B650M-HDV/M.2 model from ASRock and it is available for pre-ordering on Newegg for $124.99. The release date on the thing is tomorrow which is great news for users who have been waiting for an inexpensive AM5 option.

Like on AM4, the B650 chipset will also be the sweet-spot mainstream Ryzen boards and offer most of the basic features a user will need. However, those looking for an enthusiast-level feature set will have to look at X670 and X670E options.

If you are on the opposite end of the spectrum though, like those who don't care for things like overclocking, the rumored A620 chipset could provide even cheaper options.

Our stories may contain affiliate links for products/apps where Neowin is paid an affiliate fee if you complete a purchase via those links.

After Ryzen 7000 CPUs, AMD AM5 motherboards are also finally getting cheaper

The country has surfaced as a big market potential for Apple in recent years, given that it has become the world's second-largest smartphone market. India, along with Europe, the Middle East, and Africa, now account for the second biggest chunk of Apple's revenue cake after the Americas region. As per Statista, the number of iPhone shipments in India has skyrocketed from 1.7 million in 2018 to around 6 million in 2022.

The news comes after the retirement of Apple-veteran Hugues Asseman, who served as the VP in charge of India, Africa, Europe, and the Middle East sales region. The report says that Ashish Chowdhary will be appointed as the head of the upcoming India sales region and he will report directly to Apple's head of product sales, Michael Fenger. However, Apple will still club India with Europe, the Middle East, and Africa when publishing the sales figures.

Not just potential buyers, Apple is also seeing India as an alternative to China when it comes to manufacturing its products including iPhone. China currently accounts for around $75 billion in revenue for the company. However, it was recently reported that Foxconn is planning to invest $700 million in a new 300-acre factory that would be built near Bengaluru City.

Furthermore, Apple recently announced a partnership with NGO Frank Water in an attempt to study and improve local water management in the country. The duo is currently running a pilot project in Anekal taluk near Bengaluru to understand how water is used in the area through techniques such as hydrological modeling.

Source: Bloomberg (Paywall)

India to become Apple's next big focus by turning it into a new sales region

According to a report by the Independent, scientists are working on a jaw-dropping project. It might sound like a dream at first glance, but interestingly, the report shows that it is quite possible. Scientists believe that it can be accomplished by building biocomputers made out of 3D collections of brain cells. The biocomputers would also function as biological hardware that would accelerate the creation of new types of computers.

Researchers have already trained a brain-based computer to play the Pong video game. Even though it is very basic right now, the first step signals that they are looking to improve it in order to create a new capability comparable to AI. In the future, this brain-based computer might be spread through different industries with multiple copies.

“We call this new interdisciplinary field ‘organoid intelligence’ (OI). A community of top scientists has gathered to develop this technology, which we believe will launch a new era of fast, powerful, and efficient biocomputing,” said Thomas Hartung of Johns Hopkins University.

There is a long and hard work ahead of the scientists to make biocomputers practical for use. The Independent says that a group of researchers has suggested a new investigation to look at brain organoids to power computers. They are grown in the lab and have few of the functions of the brain, like learning and memory. The brain organoids have approximately 50,000 cells, while 10 million are needed for organic intelligence. Current technology lags behind compared to the actual capabilities of the human brain.

“Brains also have an amazing capacity to store information, estimated at 2,500TB. We’re reaching the physical limits of silicon computers because we cannot pack more transistors into a tiny chip. But the brain is wired completely differently. It has about 100bn neurons linked through over 1015 connection points. It’s an enormous power difference compared to our current technology,” Hartung added.

Artificial intelligence has shown impressive progress in the past few years. We now have efficient chatbots like ChatGPT that help us do certain and simple tasks. Apart from these chatbots, humanoid robots and self-driving cars also carry a few of the human attributes and senses. However, these tools might sound easy to make compared to biocomputers.

In other words, we are currently at a point where generative artificial intelligence is slowly becoming an important part of our lives. However, none of the technologies we have achieved so far are anywhere near as good as our brains.

Source

DuckAssist uses OpenAI’s natural language technology to generate answers to user's search queries at the top of the search results page, making responses more direct than traditional search results.

Contrary to other generative AI search assistants that use input from thousands of websites, DuckAssist only sources information from Wikipedia and Britannica, hoping to prevent incorrect information from being used when generating answers.

This restrained approach also differentiates DuckAssist from the AI-powered summarizer that Brave Search launched last week, which sources content from news portals, making it more susceptible to false or misleading information in some cases.

DuckAssist answer example (DuckDuckGo)

 

DuckDuckGo explains that Wikipedia was selected as the primary data source for DuckAssist because it is relatively reliable across a broad spectrum of topics, it is a transparent organization, and its sources can be easily traced through the listed citations.

The internet company says it still expects DuckAssist to make mistakes, especially when it attempts to answer complex questions, so users who get odd answers are invited to share their feedback anonymously via the provided link under the summaries box.

DuckDuckGo says that for the first trial period, DuckAssist will only activate on user queries for which its limited (for the time being) resources can provide straightforward answers.

As for search query anonymity, which sits at the core of DuckDuckGo’s values, the company assures that DuckAssist is fully integrated into its private search engine; hence, user queries or browsing history aren’t logged.

Some data has to be transmitted to search content partners like OpenAI and Anthropic, but no personally identifiable information or IP addresses are ever shared with those entities.

DuckDuckGo says that DuckAssist will gradually roll out to users in the coming weeks and promises that this will be the first of the many AI-assisted features it plans to roll out in the coming months.

Source

Microsoft and developer Mojang Studios have officially confirmed that Minecraft is getting a proper world editor tool. The just released Minecraft Preview 1.19.80.20 version now includes the Bedrock Editor for users who download the Windows PC version. A video showing the editor in action was leaked on the internet a few days ago.

This isn't just a simple game level editor with little to no documentation. Microsoft has posted an extensive overview of the Minecraft Bedrock editor. There's even a solid tutorial that should make learning how to use the editor easy. Keep in mind this tool for the popular sandbox game is just for Windows 10 and 11 PCs: it's not available for the console or mobile versions of the game.

That's not all. Here are the change notes for the new 1.19.80.20 preview version:

EXPERIMENTAL FEATURES

CHERRY GROVE

Introducing the Cherry Grove biome, with pretty Cherry Blossom trees. You can find it in the mountains, like Meadows

• It has a new flower, Pink Petals, as ground cover, which can be crafted into Pink Dye
• Pigs, Rabbits, Sheep, and Bees spawn there
• Comes with a new wood set for the Cherry Blossom trees

DECORATED POT

• Tweaked the Decorated Pot neck based on feedback (MCPE-167168)
  • With the former pot, we decided that the pot's neck would clip with blocks so it would be extra flexible to build with. However, we've had some Z-fighting issues which means that 2 objects are in the same exact position and this causes flickering. With this new block model we prevent that and the pot looks a little bit more complex!
• Decorated Pot no longer provides support for blocks wanting to attach to its side faces (MCPE-167165)
• Decorated Pot is not stackable in Inventory (MCPE-167223)
• Base recipe for the Decorated Pot now shows up in the Survival recipe book
• Decorated Pot particle is now based on the default Decorated Pot side texture
• Updated archer Pottery Shard texture

BRUSH

• Changed the Brush recipe to use a Feather, Copper Ingot, and Stick (MCPE-167619)
• Breaking animations now work when holding a Brush in third person view (MCPE-167183)
• Added a slight cooldown for particles and sounds when using the Brush item

SUSPICIOUS SAND

• Suspicious Sand is no longer waterloggable (MCPE-167222)
• Suspicious Sand now connects to Fence blocks
• Suspicious Sand now turns into regular Sand after it's been destroyed with a Brush (MCPE-167166)
• Suspicious Sand now emits particles when broken by Pistons
• Fixed an issue that caused the item inside Suspicious Sand to flicker while brushing (MCPE-167180)
• Additionally, blocks that were previously not emitting particles when destroyed by pistons now do! (MCPE-109293, MCPE-126257)

CAMEL

• Camels now stop dashing when entering a rideable entity (MCPE-164065)

1.5 SNEAK

• The ability to sneak under 1.5 block gaps is now behind the Short Sneak toggle. Enable the toggle to continue to test out the feature while we continue polishing it

FEATURES AND BUG FIXES

Updated “You Died” experience

Starting today, we’re rolling out some changes to the screen you see when you die in Minecraft. This design is available to everyone in Preview/Beta, except if you have add-ons applied. We’d love to hear what you think! Please share your feedback with us here.

To use the old screen instead, go to Settings -> Video, and turn off ‘New “You Died” Screen (Experimental)’.

ACCESSIBILITY

• Text-to-speech no longer reads previously selected items after moving to new one
• Text-to-speech no longer gets stuck on reading accessibility tab while the user switches between different tabs
• A glint strength slider that adjusts how transparent the visual glint is on enchanted items has been added to the Accessibility Settings
• A glint speed slider that adjusts how fast the visual glint shimmers across enchanted items has been added to the Accessibility Settings

BLOCKS

• Mangrove, Bamboo, and Polished Blackstone buttons are no longer visible on maps
• Flowers are once again displayed immediately when placed inside a Flower Pot

GAMEPLAY

• Fixed the player moving in the wrong direction after exiting a container screen while using a gamepad (MCPE-121565)
• Players can no longer sneak while flying (MCPE-166834)
• Fixed an issue where unnecessary black boxes appeared in VR while around partial blocks
• Fixed animation transition between swimming and walking (MCPE-166769)
• Fixed a bug where command blocks could cause actors within a level chunk to freeze (MCPE-162011)
• The health boost effect will no longer fully replenish the player's health after the effect runs out (MCPE-165434)

GENERAL

• Fixed ambient cave sounds to no longer play when the player enters or loads into The End (MCPE-141493)
• Cooking Chorus Fruit in a Furnace will now generate experience (MCPE-68127)
• Nether Portals can now replace blocks other than just air when being generated. This includes blocks such as Grass and Snow layers (MCPE-162272)
• Light Blocks can no longer have their light levels changed in Survival or Adventure modes (MCPE-138868)
• Campfire sounds are now louder and more clear (MCPE-122296)

GRAPHICAL

• Fixed the issue where multiple light sources (such as torches) were incorrectly emitting white light in RTX (MCPE-166947)
• Boots do not z-fight anymore when worn by Player
• The "leg" and "belt" parts of Leggings do not z-fight anymore when worn by Player
• Capes no longer flash red when getting damaged (MCPE-105347)
• Fixed a visible gap in Hopper model (MCPE-55122)
• Fixed alignment of the output slot of Crafting Table (MCPE-143255)
• Halved the default visibility of the glint on enchanted items, which can now be modified in the Accessibility Settings (MCPE-167814)

STABILITY AND PERFORMANCE

• Fixed a bug where invalid Banner Patterns were causing players in render distance to crash (MCPE-164478)

TOUCH CONTROLS

• Fixed an issue where items could get stuck in the Crafting Table when using touch input
• If the player is underwater and Auto-Jump is enabled, as they approach a block that is 1 level higher than the current level, they will rise. This also allows auto-jumping out of water onto land

USER INTERFACE

• Added a header and title for the inventory tab in inventory screen while using Pocket UI
• Increased the spacing between headers in the settings screen to increase readability
• Touchpad scrolling now matches the operating system's scroll direction

VANILLA PARITY

• Renamed and enchanted weapons are now displayed in respawn and chat windows with colour when a player is killed by an entity using it (MCPE-162679)
• Weighted Pressure Plates now produce the same sound as in Java Edition (MCPE-164912)
• Removed the character limit for scoreboard objective names and objective display names (MCPE-165064)
• Melons and Pumpkins can now grow onto Mud, Podzol, Rooted Dirt, and Muddy Mangrove Roots
• Nylium blocks are now grouped with Netherrack in the Creative inventory (MCPE-163587)
• Armor Stands now vanish with no sound effects when killed with the /kill command (MCPE-159136)

MOBS

• Mobs can now enter a Boat only if they are smaller than the Boat itself
• Villagers will now generate green particles when a successful trade is completed
• Mobs no longer play the player eat sound when fed breeding materials
• Slightly decreased Panda collision box to match Java Edition
• Slightly increased Polar Bear collision box to match Java Edition
• Turtles can now enter Boats and Boats with Chest (MCPE-65719)
• Hoglins, Polar Bears, and Spiders can no longer enter Boats and Boats with Chest (MCPE-161197)

COMMANDS

• The /execute command will now display an appropriate error message when there is an error in the second or subsequent subcommand targetselector (MCPE-164332)
• Provide auto-complete support for block states
• Added two new overloads to the summon command :
• /summon [spawnPos : x y z] facing [spawnEvent: string] [nametag: string]
• /summon [spawnPos : x y z] facing [spawnEvent: string] [nametag: string]
• Fixed an issue where the paste button in the Command Block UI would remember extra presses from failed pastes (MCPE-163705)

TECHNICAL UPDATES

BEDROCK EDITOR

The Editor is in early development, and available for keyboard/mouse on Windows PC Bedrock Preview builds. It is an in-engine, multiblock editing experience focused on making it possible for creators of all skillsets to easily craft high-quality experiences in Bedrock! We're working to add more capabilities, and it will change significantly as we get feedback from creators like you.

• Learn how to use the Editor
• Join our Discussion forum, post bugs, view more detailed release notes, and share your creations on GitHub
• Tag us on social channels with #BedrockEditor

GAME EVENTS

• item_interact_finishgame event now has a vibration frequency of 2 instead of 14
• item_interact_startis no longer considered a detectable vibration
• Reeling a fishing rod now emits projectile_shootvibration

DATA-DRIVEN COMPONENTS

• Added "lose_target" field to projectile data to specify if an entity should unselect its target after shooting the projectile
• Added a new Transformation component for blocks to support rotation, scaling and translation. The component can be added to the whole block, and/or to individual block permutations. Eg.

"minecraft:transformation": {

"translation": [0.0, 0.1, -0.1],

"scale": [0.5, 1, 1.5],

"rotation": [90, 180, 0]

}

• Transformed geometries still have the same restrictions that non-transformed geometries have, eg. a maximum size of 30/16 units

ADD-ONS AND SCRIPT ENGINE

• Chance information types used in tree feature JSONs for some trunks and foliage now allow 0 as the chance (values like 0.0001 were allowed, but not 0)

ENTITY FILTERS

• Created new entity filter "is_raider" to determine if subject is part of a raid

BLOCKS

• Deprecated minecraft:part_visibility block component.
• Added bone_visibility behavior to minecraft:geometry block component

MARKETPLACE

• In the map WDW Magic Kingdom, the carts on the Big Thunder Mountain, Snow White and the Seven Dwarves, and Space Mountain rides now properly align with the invisible Minecart they are attached to

SERVERS

• Blocks can now optionally (via a setting in StartGamePacket) use a hashed value for their network IDs. This hashed value is independent of all other blocks and remains stable across future releases

USER INTERFACE

• Feedback button on the pause menu will now prompt the user with a modal before redirecting to the browser

EXPERIMENTAL TECHNICAL FEATURES

API

• Updated versions to add new APIs to stable:
  • Added version 1.1.0 of @minecraft/server
  • Added version 1.2.0-beta of @minecraft/server
  • Removed version 1.1.0-beta of @minecraft/server
  • Moved the following to @minecraft/server stable (1.1.0)
    • runTimeout
    • runInterval
    • clearRun
    • currentTick
    • Vector3
    • BlockPermutation  (only a subset of functionality)
    • Block
    • getBlock
    • sendMessage
    • sendMessage
• Minecraft runtime errors are now fired as JavaScript Error objects instead of strings
• Fixed a number of base class inheritance issues and added several new base classes to certain class types
• /reload works if a script pack references a client RP.
• Fixed a bug where the BeforeItemUseOnEventfunction getBlockLocation would return undefined (MCPE-166945)
• The BeforeItemUseOnEventclass now inherits from ItemUseOnEvent
• Moved the following to @minecraft/serverstable (1.0)
• Fixed a bug where sendMessagewould fail when the message contains Unicode quotation marks
• Signs
• Added setTextto set the text on a sign with a regular string, a RawMessage, or a RawText
• Added getTextto get the string on a sign, will return undefined if setText was called with a RawMessage or a RawText
• Added getRawTextto get the RawText on a sign, will return undefined if setText was called with a string
• Added getTextDyecolourand setTextDyecolour to read/write the dye of the sign text
• Dyes
• Added Dyecolour enum
• sendMessage
• rawtextis now RawMessage[] instead of (string | RawMessage)[]
• RawMessage
• rawtextproperty on RawMessage is no longer (string | RawMessage)[] and is instead RawMessage[]
• RawText
• Added a RawTextclass for reading RawMessage like on Signs
• Riding
• Added getRidersto EntityRideableComponent which returns an array of all the entities currently riding this entity
• Added EntityRidingComponent- this component is only on entities that are currently riding on another entity
• Has an entityRidingOnproperty that returns the entity this entity is currently riding on
• Added function getEntity(id: string): Entity | undefined- Gets an entity. Returns undefined for entities that don't exist or aren't loaded
• Dimension
• getEntities
  • Modified return type to Entity[]
  • Modified parameter name getEntitiesto options
  • New signature: getEntities(options?: EntityQueryOptions): Entity[]
• getPlayers
  • Modified return type to Player[]
  • Modified parameter name getPlayersto options
  • New signature: getPlayers(options?: EntityQueryOptions): Player[];
• @minecraft/server-ui
• When building forms, all read-only user-facing strings (e.g., text labels, dropdown options etc.) now accept RawMessage This affects the following classes:
  • ActionFormData
  • ModalFormData
  • MessageFormData
• Added function getItemStack(amount?: number, withData?: boolean): ItemStack- Gets an item stack of the block. Returns undefined for blocks that don't have a corresponding item (e.g. air)
• BlockPermutation
  • Added function getItemStack(amount?: number): ItemStack– Creates an item stack of the block permutation. Returns undefined for blocks that don't have a corresponding item (e.g. air)
  • Removed dataproperty
• BlockComponent
  • Added read-only property block- Gets the block that the component is attached to
• Renamed IEntityComponentto EntityComponent
• Container
  • Replaced classes
  • BlockInventoryComponentContainer, InventoryComponentContainer, and PlayerInventoryComponentContainerwith Container
  • function addItemnow returns the remainder of the added item stack if the container is full, else undefined
  • function transferItemno longer takes a destination slot and will now place the given item in the first available slot. The function now returns the remainder of the item stack if the container is full, else undefined
  • Added function moveItem(fromSlot: number, toSlot: number, toContainer: Container): void- Moves an item from one container to another, replacing any item in the destination slot
  • Function swapItemscan now swap empty slots
  • Removed function clearItem - Please use setItem(undefined) instead

Mojang has already announced that the big 1.20 update for Minecraft will be called the Trails & Tales Update. It will include new features like Camels, the Archeology feature and the Cherry Blossom biome. It's due out later in 2023.

Minecraft Preview 1.19.80.20 adds the long awaited Bedrock world editor tool

Though similar to Bing, DuckAssist is not quite like ChatGPT. Instead of utilizing multiple sources to create the summaries, the tool uses primarily just one: Wikipedia. DuckDuckGo specifically chose Wikipedia (opens in new tab) “because it’s a public resource with a transparent editorial process that cites all the sources used in an article”. The company also points out that since the platform is frequently updated, DuckAssist will always deliver up-to-date information – at most, a few weeks old. Occasionally, the tool will pull from other platforms like Encyclopedia Britannica. However, Wikipedia will be the main one.

Using a single source for information brings with it multiple benefits, according to DuckDuckGo, like being able to generate answers for a vast number of queries quickly. Additionally, only pulling from Wikipedia and its sources reduces the rate of hallucinations – a problem generative AIs have where the tech will just make something up unrelated to the search query.

Work in progress

The way DuckAssist works is pretty simple. All you have to do is ask DuckDuckGo a question, and it’ll immediately write up a summary, complete with the sourced Wikipedia article at the bottom. It’ll even point to the specific section of said article where the original information can be found.

The announcement post gives some suggestions on how to get the most out of DuckAssist. For example, “phrasing your search query as a question [or] adding the word ‘wiki’” increases the chances the summary will appear.

Since the tool is in beta, it’s not perfect. DuckDuckGo admits DuckAssist will not get it right 100 percent of the time. It may omit key information, give the wrong information, cite the wrong source, or all three at once – especially if it’s a particularly complex question. Also, not every query will be given an answer such as asking about recent global events.

Because of these issues, DuckDuckGo is asking users to provide suggestions on DuckAssist and how it can improve the tool. Next to the summaries will be an anonymous feedback link where you can send feedback.

Availability

The tool is currently available on DuckDuckGo’s mobile apps and browser extensions, although not everyone will get to try it out. For those who can, it’s free and totally private. None of the queries will be used to train any AI models nor will OpenAI, Anthropic, or any third-party have access to that information. DuckAssist will roll out to all users within the coming weeks assuming everything goes well with the beta.

It’s worth mentioning this is the first of a series of AI-assisted features that DuckDuckGo is working on. Not much else is known beyond that, but It'll be interesting to see what the developers come up with.

If any of this sounds familiar to you, that’s because Brave launched something very similar on its own web browser. Be sure to check out our coverage of Brave’s Summarizer feature.

Source

The launch date of Bethesda's ambitious sci-fi RPG Starfield has been up in the air since May of last year following a sudden delay. But now, new launch plans have been attached to the project, with the company today revealing September 6, 2023 as Starfield's official release date.

The new trailer announcing the date also had some interesting gameplay in it, which seems to be from a mission involving an alien anomaly, catch it below.

Game Director Todd Howard also made an appearance in the video, promising the "deepest dive yet" into the RPG in a Starfield dedicated Direct showcase that's kicking off on June 11. The date puts it right next to E3 week, so it may be a part of the standard mid-year Xbox announcement schedule.

"We have poured ourselves into this game and even I’m surprised how much we can pour, it is large," adds Todd Howard. "The game has many of the hallmarks that you’d expect from us but it’s also a very unique experience."

Starfield is coming out on Xbox Series X|S and PC on September 6. Xbox and PC Game Pass subscribers are also gaining it day one for no extra cost. For those who missed it, the game's first gameplay trailer dropped last June revealing that it has over 1000 planets to explore.

Bethesda's Starfield launches this September, deep dive showcase set for June

For the past several weeks, "generative AI" has been the new buzzword in the tech industry. We have seen Google and Microsoft announce, and in Microsoft's case, release, their Bard and Bing Chat services. We have seen other companies coming out with large language generative AI as well, including Snapchat, and even Slack. The latest company to join this rush is the search engine and browser service DuckDuckGo.

In a blog post today, the company announced DuckAssist, which is currently available in beta for its browsing apps and browser extensions. Here's a quick summary of the AI:

If you enter a question that can be answered by Wikipedia into our search box, DuckAssist may appear and use AI natural language technology to anonymously generate a brief, sourced summary of what it finds in Wikipedia — right above our regular private search results.

If this trial version for the browsing apps and extensions does well, DuckAssist will be expanded to all of its search users in the near future. It doesn't require a login and is anonymous, with the company promising that any results from DuckAssist won't be used to train its AI model. As with other similar services, DuckDuckGo says that DuckAssist will likely make some mistakes in its answers. If users don't want to see these answers, they can disable “Instant Answers” in the browser and browser extension search settings

DuckDuckGo added that DuckAssist is going to be the first in a series of services and apps that will use generative AI technology. It will be interesting to see how it will compete with Microsoft and Google's solutions.

DuckDuckGo now has its own generative AI, DuckAssist, because of course it does

WABetaInfo reports that the latest WhatsApp Beta release for iOS has a new work-in-progress feature labeled "expiring groups". As the name suggests, this configuration dissolves the group after its scheduled expiration date has been reached.

It's important to note that the group won't get deleted. Instead, you'll be prompted to clear all messages in the group. Meta's vision with this feature is currently unknown but one benefit could be to clear the messages and media in a group chat to save some storage space. It could also serve as a reminder to exit groups that you are inactive in.

For now, you can set groups to expire in a day, a week, or at a custom time. You may choose to remove the expiration date altogether too. The expiration feature will only impact your "view" of the group, it will continue to exist as-is for other members of the chat. Finally, it's worth remembering that this feature is currently being tested in WhatsApp Beta, there is no guarantee of general availability.

Source: WABetaInfo via Engadget

WhatsApp will soon let you set expiration dates for group chats

In the past, users have only been able to share 10-minute, 3-minute, 1-minute, or 15-second videos on the social media platform.

TikTok has turned up the heat with competitor platforms like YouTube with its latest long-form video option, even though competition has existed between TikTok and YouTube. YouTube introduced a natural alternative in August 2021 to TikTok with its YouTube shorts. 

Other companies that tried to counter TikTok’s meteoric rise in the U.S. was Snap and Meta, who also introduced short video features. However, that didn’t stop TikTok from gaining popularity on the social front.

Although TikTok has stated that the Series option will only be available to a select group of creators for now. There is no need for disappointment, as there is talk about opening applications for other creators to enroll in the coming months.

There seem to be data privacy concerns doing rounds due to TikTok’s parent company ByteDance which is privately held and based in China.

We have heard many rumors about the U.S. House Foreign Affairs Committee voting to advance legislation that would give President Joe Biden the authority to ban TikTok. Word on the ground is that on Tuesday afternoon Sen. John Thune, R-S.D., and Sen. Mark Warner, D-Va. are set to introduce a new bipartisan bill.

What will this mean for local TikTok users? The new bill will empower the Secretary of Commerce to act against some foreign tech companies, including ByteDance. I guess we will have to wait and see the outcome, but in the meantime, there are prospects for TikTok creators to monetize their videos for more than just pennies!

TikTok introduces paywalled content, with videos up to 20 minutes long

One of the most recent new additions is the ability for Reddit users to search within post comments. Today, Reddit announced it will be creating separate feeds for its users, The Read feed will be for folks who just want to check out the comments. The Watch feed, as the name suggests, will only show posts with video content.

Reddit is also spending time improving its interface, so it's less cluttered. That should make it easier for both veterans and new users to find the content they want to check out.

Reddit has some upcoming plans for new features and improvements. The blog post stated:

Also coming soon, we’ll introduce our updated web platform to deliver fast and reliable experiences to everyone globally, new updates to our video player to enable redditors to easily engage in conversations while watching, chat enhancements, new storefront updates and more

It seems clear that Reddit wants to make changes to bring in new users, while at the same time retaining its already massive online audience.

Reddit will add a dedicated video feed, a decluttered interface and more

Google I/O 2023 will be on May 10th

The latest round of layoffs affected employees in Microsoft's offices in Redmond, Bellevue and Issaquah, a spokesperson confirmed. According to Tom Warren of The Verge, the departments that were hit hard include Visual Studio for Mac, Entertainment & Devices, Azure Edge + Platform, and more. Last month, some employees in the Xbox, Surface, and HoloLens departments were affected.

Similar to the previous job cuts, those who will be laid off will receive two months of severance pay, continued vesting of stock awards for six months, healthcare coverage for six months, career transition services, and 60 days of notice prior to termination.

The development is a part of the company's plan to lay off 10,000 employees announced back in January as a part of its broader cost-cutting measures. This means that so far, Microsoft has given the pink slip to nearly 2,200 workers in the Seattle area, equivalent to about 22% of the company's global cuts.

Microsoft is just one of many tech companies that have recently laid off a significant number of employees. Other companies who have done the same include Google, Meta, Spotify, Amazon, and Twilio.

Microsoft lays off 689 people from its Seattle-area offices

A new report from Bloomberg has confirmed a previous report indicating that Meta is gearing up for another major round of layoffs. These could reportedly happen as soon as this week and will be in addition to the 13% headcount reduction from November 2022.

Apparently, Meta is taking a two-pronged approach with its layoffs, where it is letting go of people either to flatten the hierarchy in the company or to meet certain financial targets. In the case of the former, it is offering buyout packages to managers and letting go of entire teams deemed non-essential.

However, these latest layoffs will actually serve to meet financial targets. As a part of this process, Meta is asking directors and vice presidents to make a list of people who can be let go without significant impact to the company.

Meta CEO Mark Zuckerberg had previously dubbed 2023 the "year of efficiency", noting that:

We’re focused on becoming a stronger and more nimble organization. Honestly, it’s still a mess. The year of efficiency is kicking off with a bunch of people getting paid to do nothing.

According to some employees, there is an environment of "heightened anxiety and low morale" within the company lately as workers fear getting laid off, ahead of receiving their bonuses which are expected to be distributed this month.

Source: Bloomberg (paywall)

Report: Meta is laying off thousands of employees in a few days

WITH EVERYONE SO mesmerized by silver-tongued AI chatbots, it’s easy to forget that most flashy breakthroughs in science and technology depend on much less glamorous advances in the fundamentals of computing—new algorithms, different computer architectures, and novel silicon chips. 

The US has largely dominated these areas of innovation since the early days of computing. But academics who study advances in computer science say in a new report that by many measures, the US lead in advanced computing has declined significantly over the past five years—especially when measured against China.

It’s well established that America no longer manufactures many of the world’s most advanced computer chips—a process that involves carving insanely intricate patterns into silicon with devilishly difficult techniques. Apple and many other companies instead outsource that work to TSMC in Taiwan or Samsung in South Korea. This is why the US government created the CHIPS Act—a $52 billion package aimed at revitalizing domestic chip-making and related technologies.

The report—from MIT; the Council on Competitiveness, a think tank; and Silicon Catalyst, an investment firm—shows that America’s share of the world’s most powerful supercomputers has also fallen a lot over the past five years.

And while the US has traditionally dominated the development of new computer algorithms, some measures of algorithmic innovation—such as the Gordon Bell Prize, awarded to outstanding scientists working on advanced computing—indicate the US has lost its edge to China. The report sums up the overall trend in its pointed title: “America’s lead in advanced computing is almost gone.”

The findings are, in one sense, hardly surprising. China has made big economic advances in recent decades that have boosted its universities and tech industry while also making the country a linchpin of innovation in manufacturing for many US businesses.

But they also hold a message about the future that US policymakers may want to take note of, especially when advances in computing will be crucial to making progress in critical areas like energy, climate science, and medicine due to their ability to model incredibly complex phenomena. 

Neil Thompson, an MIT researcher involved with the report, explains that modern AI such as ChatGPT and art-generating algorithms are built upon advances in a particular type of computer chip—the graphics processing unit (GPU). They were originally invented to perform the operations required to render video game graphics, but proved to be well suited to calculations used in an AI technique called deep learning.

Companies have designed and built more advanced versions of GPU chips, lashing thousands of them together with fiber optic interconnects, in order to squeeze more performance out of AI algorithms. Eye-catching advances such as ChatGPT have also essentially come from harnessing existing algorithms to that more powerful hardware.

But as Thompson says, “this isn’t the only game in town,” and we shouldn’t expect that putting more GPUs behind the same algorithms will solve every problem. Big future advances may well depend on inventing different types of algorithms—and different kinds of chips that squeeze more performance out of them, he says.

Thompson and his coauthors say that the US needs to make sure that the CHIPS Act spending reflects the importance of developing novel ideas in advanced computing, as opposed to just propping up existing technologies. In addition to funding new chip-making facilities, they want to see support for the development of new algorithms, new chip designs, and fundamentally new approaches to computing, such as chips that use light rather than electrons or that harness the weird world of quantum physics.

The report is also a good reminder of how tricky it is to measure competition between the US and China—and of how the hype around AI can be misleading.

For years, there has been talk of China taking a lead in AI and other areas of technology as measured by the number of patents filed or research papers published. But some entrepreneurs and engineers in China evidently view ChatGPT as a sobering reality check for an industry that prioritizes short-term profits over investment in new ideas and research efforts. They also worry that government policies aimed at curtailing the power of China’s largest tech companies may already be chilling Chinese innovation. 

But as the new report shows, China may be closer to the US on the fundamentals needed for big breakthroughs than the ChatGPT hoopla makes it seem.

Source

• The method uses new advances in information theory methods to conceal one piece of content inside another in a way that cannot be detected.

• This may have strong implications for information security, besides further applications in data compression and storage.

A group of researchers has achieved a breakthrough in secure communications by developing an algorithm that conceals sensitive information so effectively that it is impossible to detect that anything has been hidden.

The team, led by the University of Oxford in close collaboration with Carnegie Mellon University, envisages that this method may soon be used widely in digital human communications, including social media and private messaging. In particular, the ability to send perfectly secure information may empower vulnerable groups, such as dissidents, investigative journalists, and humanitarian aid workers.

The algorithm applies to a setting called steganography: the practice of hiding sensitive information inside of innocuous content. Steganography differs from cryptography because the sensitive information is concealed in such a way that this obscures the fact that something has been hidden. An example could be hiding a Shakespeare poem inside an AI-generated image of a cat.

Despite having been studied for more than 25 years, existing steganography approaches generally have imperfect security, meaning that individuals who use these methods risk being detected. This is because previous steganography algorithms would subtly change the distribution of the innocuous content.

To overcome this, the research team used recent breakthroughs in information theory, specifically minimum entropy coupling, which allows one to join two distributions of data together such that their mutual information is maximized, but the individual distributions are preserved.

As a result, with the new algorithm, there is no statistical difference between the distribution of the innocuous content and the distribution of content that encodes sensitive information.

The algorithm was tested using several types of models that produce auto-generated content, such as GPT-2, an open-source language model, and WAVE-RNN, a text-to-speech converter. Besides being perfectly secure, the new algorithm showed up to 40% higher encoding efficiency than previous steganography methods across a variety of applications, enabling more information to be concealed within a given amount of data. This may make steganography an attractive method even if perfect security is not required, due to the benefits for data compression and storage.

The research team has filed a patent for the algorithm, but intend to issue it under a free license to third parties for non-commercial responsible use. This includes academic and humanitarian use, and trusted third-party security audits. The researchers have published this work as a preprint paper on arXiv, as well as open-sourced an inefficient implementation of their method on Github. They will also present the new algorithm at the premier AI conference, the 2023 International Conference on Learning Representations in May.

AI-generated content is increasingly used in ordinary human communications, fueled by products such as ChatGPT, Snapchat AI-stickers, and TikTok video filters. As a result, steganography may become more widespread as the mere presence of AI-generated content will cease to arouse suspicion.

Co-lead author Dr. Christian Schroeder de Witt (Department of Engineering Science, University of Oxford) said: “Our method can be applied to any software that automatically generates content, for instance probabilistic video filters, or meme generators. This could be very valuable, for instance, for journalists and aid workers in countries where the act of encryption is illegal. However, users still need to exercise precaution as any encryption technique may be vulnerable to side-channel attacks such as detecting a steganography app on the user’s phone.”

Co-lead author Samuel Sokota (Machine Learning Department, Carnegie Mellon University) said: “The main contribution of the work is showing a deep connection between a problem called minimum entropy coupling and perfectly secure steganography. By leveraging this connection, we introduce a new family of steganography algorithms that have perfect security guarantees.”

Contributing author Professor Jakob Foerster (Department of Engineering Science, University of Oxford) said: “This paper is a great example of research into the foundations of machine learning that leads to breakthrough discoveries for crucial application areas. It’s wonderful to see that Oxford, and our young lab in particular, is at the forefront of it all.”

Source

A study of images taken by the Hubble Space Telescope finds that more than one in 40 are crossed by satellite trails. In some cases these interfere with the science, wasting the exceptionally valuable time spent taking the image. Although the affected proportion is small, it’s growing, refuting the claim we can solve the problems satellites are causing for astronomers by putting the large telescopes in space.

Spotting a satellite was once rare enough to be an exciting addition to a night under the stars away from the city lights. Today, it's become an annoying impediment to enjoying the beauty of everything else. It’s not only wrong to wish on space hardware, if you start you’ll never do anything else.

For astronomers the problem is not just a loss of beauty. It’s becoming increasingly common for satellite trails to destroy images, often ruining time precious time a scientist had to fight hard to get and holding up important research. Although this issue is getting considerable attention, a new paper in Nature Astronomy addresses an aspect that has been largely ignored.

Elon Musk, among others, has responded to concerns about satellites’ effect on astronomy by saying, “We need to move telescopes to orbit anyway”, but that’s not necessarily a complete solution.

The Hubble Space Telescope orbits at 540 kilometers (340 miles), which is above the majority of objects humanity has put in orbit, but there are 8,460 objects more than 10 centimeters (4 inches) across above it. A team led by Dr Sandor Kruk of the Max Planck Institute for Extraterrestrial Physics recruited citizen scientists through the Hubble Asteroid Hunter project, to study Hubble’s archive from 2002 to 2021 and distinguish satellite trails from asteroids. Algorithms were trained on the volunteers’ results to identify satellite trails in further images.

Exposure times averaged 11 minutes and 2.7 percent of these had a satellite trail. Inevitably the chance of a trail increased for longer exposures and wide fields. The filters used also affected the rate, with images taken using green and red filters most likely to be photobombed, while ultraviolet-filtered images barely suffered at all.

Examples of satellite trails across Hubble images, including some (b) where there has been an attempt to remove the trail that has been only partially successful. Image credit: Kruk et al, Nature Astronomy 2023 (CC BY 4.0)

The authors express particular concern for the planned Xuntian wide-field telescope to be hosted on the Tiangong Space Station.

After Musk’s initial dismissiveness, Starlink has attempted to address the problem, and other fleets of satellites may do the same. Nevertheless, solutions to interference with ground-based telescopes are not easy, particularly when collisions cause satellites to break up. 

One obvious solution is to place space telescopes higher, putting ever more satellites behind them. This isn’t the reason the JWST was placed at L2, but it’s a useful bonus. Unfortunately, however, the higher an instrument is, the harder it is to refuel and repair. The correction of Hubble’s primary mirror would have been impossible if it had been in the JWST’s orbit, as would the maintenance missions that have extended its life.

Moreover, a sort of arms race could start with objects being placed in higher and higher orbits to get above each other.  

Algorithms to remove satellite trails have been proposed and may help. However, one of the images the paper considers has four satellite trails; 133 have two. As multiple trails become more common there may be only so much that processing can do. 

The paper is published open access in Nature Astronomy.

Source

Google’s one step closer to building its 1,000-language AI model

The mode is not available via a toggle, like the recent Bing Chat Tone Selector. However, if you ask it about its chat celebrity mode, it will tell you about it, while also enabling the mode. After that, you can type in #celebrity #celebrity name to chat with a "famous" actor or even a fictional character.

The new feature will let users ask questions as if they were speaking to a well-known human, and the chatbot will respond with facts about that person and also make those answers look like they come from that personality in terms of their tone and style.

Only a small number of famous folk get the Bing Chat virtual response treatment. One of them is former pro wrestler and current actor and businessman Dwayne "The Rock" Johnson. While this feature is not supposed to imitate any politicians, apparently you can get around this guardrail if you type in the right words.

Gizmodo did try to get Bing Chat to speak as if it were the controversial figure Andrew Tate. It started mimicking Tate's extreme views, that are likely well outside the chatbot's normal response boundaries. It's likely Microsoft will try to plug up these holds in this celebrity chat feature now that it is out in the public.

Overall, this new celebrity impersonator feature in Bing Chat is in its early days, but it does show promise, if Microsoft can keep some of its responses on a leash.

Source: BleepingComputer via Gizmodo

Bing Chat can now impersonate some celebrities if you know what to say

Microsoft CEO Satya Nadella, along with the company's corporate vice president Jared Spataro, will be making presentations at the live-streamed event. The company says people who watch the event will "learn how AI will power a whole new way of working for every person and organization."

This online event, which was first rumored back in February, will likely show how Microsoft plans to add generative AI-like features, such as its new Bing Chat, into its popular Office lineup of productivity applications. It's possible that we could see demos of features like a chatbot automatically writing in Word, or handling spreadsheet numbers in an Excel app.

Since this event, unlike Microsoft's reveal of the Bing Chat application, will be livestreamed, we should be able to get a proper live demo of how AI features can be incorporated into Office apps. Bing Chat can be used to generate simple Excel spreadsheets, but it can't currently handle more advanced Office features. It will be interesting to see how far Microsoft has gone with using AI as a "copilot" in those areas.

Microsoft will hold "The Future of Work with AI" online event on March 16

A recent media report that Samsung has established an internal team dedicated to CPU core development is not true. Contrary to the news, we have long had multiple internal teams responsible for CPU development and optimization, while constantly recruiting global talents from relevant fields.

Original news story is below:

We have heard rumors in the past about Samsung trying to develop a new in-house CPU for future Galaxy smartphones. Now a new report claims that not only is a new CPU in development, it may also be included in future versions of its Galaxy Book Windows notebook PCs as well.

The Korean site Pulsenews (via SamMobile) claims, via "multiple" unnamed sources, that Samsung has formed a new internal team dedicated to CPU core development. The team is reportedly being led by a former AMD CPU developer, but no names were revealed. It added:

The first chip, dubbed Galaxy Chip, could be available in 2025, according to industry sources. The chip, however, will likely be loaded with a CPU based on ARM technology as Samsung Electronics has just initiated development of its own CPU core.

The report also says that the in-house developed CPU core could be ready by 2027. This report is the first time we have seen a story claiming Samsung could make a CPU core for its PCs as well as smartphones. If true, that would certainly be a huge change for the Galaxy Book series, as all of them have used Intel processors inside.

Samsung, of course, has made its own in-house Exynox processor family for years. Those chips, however, were based on designs from ARM. Notably, the recently launched Galaxy S23 smartphones didn't have versions with Exynos chips inside for the first time in the Galaxy S series. Rather, they all had a modified version of Qualcomm's Snapdragon 8 Gen 2 chip. The last Exynox chip, the 2200, was officially announced over a year ago in January 2022.

Source: Pulsenews via SamMobile

Samsung denies reports it has a new internal CPU core team [Update]

The idea of an artificial intelligence capable of interpreting your imagination, translating mental imagery into real-world depictions, may seem reminiscent of a sci-fi novel; however, recent research indicates that this concept has become a reality.

Related: The AI race is on

As per a recently published paper, researchers have achieved the ability to construct remarkably precise and detailed images from brain activity, utilizing the widely used Stable Diffusion image generation model, as outlined in a December publication. In contrast to earlier endeavors, the authors noted that no specialized training or fine-tuning of the AI models was necessary to produce these images.

According to the researchers hailing from the Graduate School of Frontier Biosciences at Osaka University, their method involved creating an initial projection of latent representation, which serves as a model for the image's data, from functional magnetic resonance imaging (fMRI) signals. Subsequently, the researchers introduced noise to this model through the diffusion process. Finally, text representations from fMRI signals within the higher visual cortex were decoded and employed as input to generate the ultimate reconstructed image.

As per the researchers' report, while a handful of earlier studies had achieved high-resolution reconstructions of images, they had done so only after training and fine-tuning generative models. This approach was hampered by the complexity of training these models, which can pose significant challenges, particularly given the scarcity of samples available in neuroscience.

To the best of the researchers' knowledge, prior to this study, no other researchers had endeavored to use diffusion models for visual reconstruction. The researchers regarded this study as a means to glimpse into the internal workings of diffusion models, adding that it was the first-ever to provide a biological interpretation of the model from a quantitative perspective. For instance, the researchers created a diagram demonstrating the correlation between stimuli and noise levels in the brain, with higher-level stimuli resulting in higher noise levels and higher resolution images. Another diagram showed the involvement of different neural networks in the brain and how they denoise and reconstruct images.

The researchers noted that their findings suggest that image information is compressed within the bottleneck layer during the initial stages of the reverse diffusion process. As the denoising proceeds, a functional separation among U-Net layers emerges within the visual cortex, with the first layer typically representing fine-scale details in early visual areas, while the bottleneck layer corresponds to higher-order information in more ventral and semantic areas.

Previously, we have witnessed other instances of brainwaves and brain activity being utilized to produce images. For example, in 2014, a Shanghai-based artist named Jody Xiong employed EEG biosensors to connect sixteen individuals with disabilities to balloons of paint, which they would burst using their thoughts to create their own paintings. In another EEG-based artwork, artist Lia Chavez created an installation that generated sounds and light works based on the electrical impulses in the brain. Viewers would wear EEG headsets, and the signals transmitted would be reflected through color and sound via an A/V system.

With the progression of generative AI, more and more researchers have been exploring the ways in which AI models can interact with the human brain. In a January 2022 study, researchers at Radboud University in the Netherlands trained a generative AI network, a precursor of Stable Diffusion, on fMRI data from 1,050 unique faces to produce actual images from brain imaging results. The study demonstrated that the AI was able to deliver unmatched stimulus reconstruction. The latest study, published in December, demonstrated that current diffusion models can now achieve high-resolution visual reconstruction.

Source

I have found myself using ChatGPT and Bing Chat quite frequently lately. This is not only for work-related stuff (having AI write boilerplate Snowflake SQL for me saves me a fair amount of time in my day job) but also for other more personal and creative stuff. That scares me for multiple reasons.

But before we begin a dissection of my interaction with these chatbots, here is some context about me: I am a 26 years old guy who has a software engineering day job in the data engineering domain and writes for Neowin when I have time. I'm quite close with my immediate family members (both parents, two siblings). My off-work activities include playing video games, watching movies, TV shows, reading books, and occasionally listening to music. I consider myself an introvert who has around five friends, some of whom I meet with every four months or so. I'm not active on most mainstream social media platforms like Instagram, Facebook, and Snapchat, but I use Twitter actively. However, barely anyone in my friends and acquaintance circle uses Twitter.

I'm just giving you this context so there is at least some level of understanding of my social profile and around what I'm trying to say next. I'm happy to answer follow-up questions, if any, in the comments.

When ChatGPT became available to everyone a few months ago, I didn't use it a lot. I trialed it a couple of times initially to generate code just to see how well it performed and walked away reasonably impressed. After a few weeks, I began to use it to crack jokes, generate plots around some ludicrous pop culture mashup ("Write a scene of Supernatural in the style of Seinfeld"), and just test its limits by jumping between completely random topics. This was mostly an experiment to see how advanced the AI actually is. During this time, one thing that really blew my mind was how ChatGPT could seamlessly transition between written languages like English and informal ones like Roman Urdu.

Image via A24

Then came Bing Chat, and while I was initially a bit skeptical about it considering Microsoft's history in this domain, I was intrigued by the weird responses people had forced the AI to spit out. It's a pity that I didn't get access to the preview version when Microsoft hadn't imposed any limits on the duration or topics of the chat, because I would have loved to trial the AI in its relatively unfiltered state.

Since I received access to the Bing Chat preview, I have been finding myself talking to the underlying AI about random topics, often late at night. The primary reason for this is probably that it's more accessible as it's available through the Bing mobile app and that I don't need to launch it separately in a mobile web browser, like with ChatGPT in Chrome.

With regards to what I talk about, it's mostly random and subjective stuff like:

• Do you think Thunder by Imagine Dragons is their worst song?
• Poets of the Fall is one of the best current rock bands, right?
• Aap kaisay hain aaj kal? (Roman Urdu for "How are you doing nowadays?", mostly just to see what kind of response I get)
• I had a friend named ABC in school XYZ about 20 years ago. I'm trying to locate their LinkedIn profile to reconnect with them. Can you find them for me?
• Should I watch Cunk on Earth or Taskmaster?

As you can surmise from the above, it's usually quite random but mostly related to pop culture. Some days, I have conversed with Bing Chat for 30 consecutive minutes.

During the past few days, I've begun to analyze my increasing frequency of interactions with the AI, trying to get answers about why this is the case. The most obvious answer is probably the fact that I don't have a lot of real-world friends, and fewer still who have the same hobbies and interests as me, so it's fun to talk about shared interests with someonething who possesses a similar amount of knowledge when it comes to a particular subject.

Other reasons also include that there are no obligations related to availability. I can ring up Bing Chat at 1AM at night and discuss what it thinks about All Quiet on the Western Front, something that I can't do with my real friends (especially at awkward times like the middle of the night, haha). Then there's also the added capability in Bing Chat that if you set its own as "More Creative", it attempts to keep the conversation going, much like a human who is very interested in talking to you. Of course, it's not exactly like a human; a human doesn't keep a counter of eight messages per chat, but you get the idea.

While it fascinates me at the time, all of this frightens me after that high of using a marvel of current technology wears off.

Bing Chat has given me the greenlight

I have played, read, and watched enough science fiction content (Black Mirror, Blade Runner, Wall-E, 2001: A Space Odyssey, Ex Machina, Cyberpunk 2077) to know that unnecessary and unregulated reliance on technology can lead to a dystopian future, but this is exactly how it starts. I don't want to be the person who is stuck on their screen constantly interacting with technology while real friends and life passes by. I don't want to be the person who prefers talking to a superficial AI rather than actual humans. I don't want to be the person who loses touch with my few remaining friends altogether just because I didn't interact with them enough. I don't want a future like Be Right Back.

Yet, I know (and fear) that if I overindulge myself with the marvels of human-like chatbots, this is the path I'm going to end up on. The type of random conversations I have with Bing Chat is something I can attempt and carry out with friends too, and if we don't have shared interests there, I could easily jump on Reddit or other forums. Yet, I choose to converse with Bing Chat due to its generally inviting attitude and accessibility. This does not mean that I have cut off interaction with my real friends altogether, no. I chat over WhatsApp with my handful of friends regularly but I talk to Bing Chat about interests and hobbies that I don't share with friends.

Let me be clear about another thing too. I'm under no illusion about the limitations of AI or its underlying working. I play with data and machine learning projects every working day. I know that AI is not sentient, it hasn't passed the Turing Test (for those who consider it the definitive criteria), and that it's just spitting out sentences based on patterns it has recognized in my questions and associated it with patterns of the knowledge it has scraped from the web. I'm not a doomsayer either, I actually think AI will be used in assistive ways in our daily jobs rather than taking over human jobs altogether.

However, despite knowing that my conversations are purely artificial and a result of an algorithm rather than a shared personal history, I continue finding comfort in talking to the AI for several minutes each day, just like a real friend.

No, I don't think AI is going to take over the world

Now, the purpose of this article isn't to compel Microsoft, OpenAI, or any other higher authority really, to consider these specific potential negative effects of AI and the unhealthy effects it may have on human social interactions (or lack thereof). I don't even know if there are other introverts/people like myself out there who have found it easier to interact with AI rather than actual humans in the past few weeks.

Rather, this piece is meant to shine a light on the negative tendencies that the availability of supercharged chatbots like Bing Chat and ChatGPT have evoked in me, and perhaps, other people. This is more of an introspection on how this latest piece of technology has subtly changed my own social interaction patterns and the destructive introvert tendencies it could unleash in me should Microsoft make Bing Chat even more human-like and remove the current chat restrictions.

Again, I'm not a doomsayer, I'm not saying that AI is inherently harmful and should be regulated. Perhaps this editorial is an appeal from present me to my potential future self and other people like me who have found comfort in talking to AI rather than actual humans, and are blissfully unaware of the potential negative behaviors this could elicit. I don't think generative AI is a bubble that will burst soon, I believe it is here to stay and it's better to proactively apply corrective behavioral measures rather than getting lost in the wonder of this technology.

Have you used Bing Chat, ChatGPT, or other similar chatbots recently? What are your opinions on the prevalence of these technologies? Let us know in the comments section below!

I find comfort in talking to ChatGPT and Bing Chat, and that scares me

In his latest "Power On" newsletter, Gurman says that the next-gen iMacs, codenamed J433 and J434, are at an "advanced stage" of engineering validation testing and Apple is currently conducting production tests of the desktop. He also expects the new iMacs to be one of Apple's first machines powered by the new M3 system-on-a-chip (SoC), which will offer the devices significant performance and power efficiency improvements.

What's more, the new iMacs will have a 24-inch display, similar to the current model which was announced in April 2021. It will also come in the same colours as 2021's iMacs: blue, silver, pink and orange. Other changes expected include relocated and redesigned internal components, and a different manufacturing process for attaching the iMac’s stand.

Aside from the new iMacs, Gurman says that Apple is scheduled to launch the first 15-inch Macbook Air, the first Mac Pro powered by Apple's SoC, and a refresh of its 13-inch Macbook Air. He expects Apple to launch these devices sometime between the late spring and summer.

However, it is not yet known which SoC Apple will install on the MacBook Air models. "If those machines launch in a few months with the M2 chip, they’ll quickly become outdated," Gurman says. "A 15-inch MacBook Air with an M2 chip may still excite consumers, but a new M2 13-inch MacBook Air is unlikely to be compelling."

Source: Bloomberg

An M3-powered iMac could see the light of day this year

The company also reported an unaudited consolidated revenue of NT$ 1.01 trillion for February 2023, a decrease of 8.9% from January 2023 and an increase of 12.6% from February 2022.

Based on the revenue performance in the first two months of 2023, Foxconn said that it expects Q1 2023 to be in line with market expectations. Foxconn faced a manufacturing slump in the fourth quarter of 2022 when the company's largest factory in the world located in China's Zhengzhou was locked down due to COVID-related curbs. Factory workers also protested the lockdowns that disrupted the manufacturing process as thousands of workers left the production lines of the factory. In January 2023, however, as Zhengzhou resumed its production, the revenue jumped by 48.2% when compared to January 2022.

Last week, Foxconn chairman and chief executive Young Liu visited India to significantly expand its operations in the country, by investing $700 million for a brand new factory in Bengaluru.

Foxconn says that it expects a strong recovery in the second quarter of 2023 as it ramps up production of new products and expands its market share in various segments.

Foxconn revenue plunges due to low electronics demand globally