Tag: artificial-intelligence

Apple Event Announced; Google Gemini AI Image Model Upgrade; AT&T Picks Up Spectrum from EchoStar; Farmers’ Insurance Data Breach

August 26th, 2025

Apple has put out the invite to their next event, with the tease ‘Awe Dropping.’ The rollout of the new iPhones will be Tuesday, September 9th at 10 AM Pacific at the Steve Jobs Theater at Apple Park in Cupertino. Macrumors.com reports that we will see the expected thin profile iPhone 17 Air, in addition to the iPhone 17, the 17 Pro, and 17 Pro Max. Also expected…the Apple Watch Series 11, and a further refreshed Watch Ultra 3, and the Watch SE. Don’t be surprised if there is an updated AirPods Pro 3 as well, and maybe a new HomePod mini. We’ll have a full recap here, of course.

Google is upgrading its Gemini chatbot with a new AI image model that is supposed to allow finer control for users over editing pics…a move intended to catch up with the popular ChatGPT image tools from OpenAI. The more precise edits to images require the usual natural language requests from the users, and are aimed at preserving the consistency of faces, animals, and the like. Google is hoping to close the gap with ChatGPT, which gets over 700 million weekly users. Right now, google Gemini has 450 million MONTHLY users.

AT&T is picking up $23 billion worth of spectrum licenses from EchoStar, the parent of Dish Network, Sling TV, and Boost Mobile. EchoStar had been under the gun from the FCC for not building out anything on the spectrum. The FCC has a ‘use it or lose it’ policy on spectrum licenses. The licenses cover over 400 markets in the US, and AT&T says it plans to start work as soon as possible. The licenses are for low-band and mid-band spectrum, which are used for 5G and LTE networks.

Farmers Insurance says it was hit by a data breach that exposed information of over a million policy holders. Mashable.com reports that the third party attack came via a vishing scam that was related to a series of SalesForce associated attacks. According to Farmers, the compromised data includes names, addresses, dates of birth, driver’s license numbers, and in some cases, the last four digits of Social Security numbers. Reports indicate that around 1,111,386 people were affected across 10 states: California, Washington D.C., Iowa, Maryland, Massachusetts, New York, New Mexico, North Carolina, Oregon, and Rhode Island.

I’m Clark Reid and you’re ‘Technified’ for now.
iPhone Fold Next Year-5 Cams; xAI (Musk) Sues Apple and OpenAI; YouTube Secretly Edited Videos With AI, Waymo Can Test Self-Drivers in NYC

August 25th, 2025

Even though the new iPhones are due out in just weeks, including the anticipated iPhone Air…or whatever they actually end up calling the new ‘skinny’ iPhone, tidbits are already leaking out about next year’s Folding iPhone. 9to5mac.com reports that the Cupertino folder will feature 5 cameras, and will have Touch ID…and will use and Apple cellular modem. A number of these new details come from Bloomberg’s Mark Gurman and his Power On newsletter. As already reported, the iPhone folder will use the book style, not the flip phone style. Expect one cam on the front screen, one on the inside, and two on the back. The phone will use Touch ID, and will run on Apple’s in-house designed cellular modem…as Apple switches its entire line to the new modems. The camera setup will allow the phone to be used much like a regular iPhone when closed…with a selfie cam on one side and the two main cams on the back. When you open the phone up, the inside cam takes over as the selfie cam. Why no Face ID? In a word, space. Touch ID doesn’t require as much room, and since the folding phone will be notably thicker than a regular iPhone, every millimeter counts.

Elon Musk’s xAI is suing Apple and OpenAI, claiming that their deal to build ChatGPT into the iPhone is stifling competition in the AI industry. According to theverge.com, Musk-owened X Corp, the parent of xAI, accuses Apples App Store of “deprioritizing” rival chatbots and “super” apps, including Grok and X. Musk’s companies claim that iPhone users “have no reason” to download third-party AI apps because the company “force[s]” users to use ChatGPT as their default chatbot app when enabling Apple Intelligence. “Apple and OpenAI have locked up markets to maintain their monopolies and prevent innovators like X and xAI from competing,” the companies allege. “This latest filing is consistent with Mr Musk’s ongoing pattern of harassment,” OpenAI spokesperson Kayla Wood said in a statement. As for Apple, it had already put out a statement saying that the App Store is “designed to be fair and free of bias.”

YouTube has, in recent months, secretly used AI to tweak some creators’ videos without letting them know or asking permission. Bbc.com notes that one content creator noticed that his hair was different and that he seemed to be wearing makeup. In another case, wrinkles in a shirt were more sharply defined. Some ears appeared to be warped. YouTube has finally at least come clean on admitting the AI changes. Rene Ritchie, YouTube’s head of editorial and creator liaison, posted on “X: We’re running an experiment on select YouTube Shorts that uses traditional machine learning technology to unblur, denoise, and improve clarity in videos during processing (similar to what a modern smartphone does when you record a video). YouTube is always working on ways to provide the best video quality and experience possible, and will continue to take creator and viewer feedback into consideration as we iterate and improve on these features.” YouTube hasn’t answered media questions as to if it will now seek user permission before using AI to tweak their videos. I, for one, would really rather they didn’t use AI to give me Vulcan pointy ears like Mr Spock!

Waymo can now go forward and test its self-driving cars in New York City. The word came from the office of Mayor Eric Adams. Engadget.com said the company has a permit to operate the autonomous vehicles in parts of Manhattan and Downtown Brooklyn. Right now, a small fleet of 8 vehicles will be able to operate until late September 2025. At that point, Waymo will be offered the opportunity to get an extension if all goes well. Right now, human operators will be on board…New York state law prohibits operation of vehicles without a driver behind the wheel. Waymo is lobbying to get this regulation changed.

I’m Clark Reid and you’re ‘Technified’ for now.
OpenAI Returns GPT-4o Over User Revolt; Amazon Same-Day Grocery Delivery-2300 Cities by Year’s End; TeaOnHer-Driver Licenses Exposed in 10 Minutes; US Secretly Tracked Diversion of Dell & Super Micro Chips to China

August 13th, 2025

OpenAI has had a pretty rocky rollout of its latest-greatest…or allegedly greatest, GPT-5. in fact, it has been rough enough that arstechnica.com reports that the user backlash has forced OpenAI to let users revert to GPT-4o. That AI model now appears in the model picker for all paid ChatGPT users by default (including ChatGPT Plus accounts), marking a swift reversal after thousands of users complained about losing access to their preferred models. The backlash wasn’t pretty at all, with a Reddit thread titled “GPT-5 is horrible” amassing over 2,000 comments in just days. OpenAI has now made some modifications to address the user outrage over GPT-5. Rate limits for GPT-5 Thinking mode increased from 200 to 3,000 messages per week, with additional capacity available through “GPT-5 Thinking mini” after reaching that limit. The company also added new routing options—”Auto,” “Fast,” and “Thinking”—giving users more control over which GPT-5 variant handles their queries.

Amazon has announced that it will offer same-day grocery delivery in some 2300 cities by the end of the year…that’s double what they are doing now. According to Bloomberg, users will be able to order perishables like produce, meat, seafood, dairy, and baked goods…in addition to frozen foods and household items. Same day delivery for grocery items is free for Amazon Prime subscribers on orders over $25 in most cities. For non-members, it runs $12.99 per order, no matter what the order size is.

We recently reported on the Tea app, an app that lets women clue each other in on men they know or have dated, and tip them off to ‘red flags’ with those men. It turned out, the app was leaking the women’s information, including driver’s licenses. Well, as the old saying goes, ‘what’s good for the goose is good for the gander,’ only in this case it’s what’s bad! Techcrunch.com says the app for guys, TeaOnHer, also leaked like a sieve…with driver’s licenses found by TechCrunch within 10 minutes of being sent a link to the app in the App Store! The flaws that allowed this kind of leak have apparently been resolved now, but not before thousands of guys’ driver’s license info was exposed. Once again, I get to warn…no matter what an app might potentially do for you…DON’T upload your driver’s license or government ID to use it! With new laws in the United Kingdom, this has become a major issue, with authorities there enacting a law requiring sites to demand age verification…even Wikipedia! The Wiki is fighting it, and may even make itself unavailable in Britain over this issue if it loses in court.

The US government has apparently put location tracking devices in targeted shipments of advanced AI chips considered at hight risk of illegal diversion to China. Benzinga.com reports that the trackers are hidden in server packaging from Dell and Super Micro Computer for units containing chips from NVIDIA and AMD. Some shipments of the restricted chips have resumed, after the chipmakers agreed to virtual extortion by the Trump administration of 15% of their profits on sales of the chips in China. Normally the Commerce Department’s Bureau of Industry and Security oversees such operations, but this appears to also involved Homeland Security Investigations, and the FBI. Super Micro declined to comment on the issue, and Dell said it was ‘not aware’ of such trackers in its shipments. To me this has the vibe of Mad Magazine’s old ‘Spy vs. Spy’ cartoons. Will the Black Spy get the upper hand on the White Spy…or will the Lady in Gray bamboozle both of them?

I’m Clark Reid and you’re ‘Technified’ for now.
Live Translation Coming to AirPods; Musk Rages Apple Favors OpenAI over Grok; Ford Universal EV Platform Called Game Changer; Reddit Blocks Internet Archive Over AI Scraping

August 12th, 2025

Some folks have noted new images in iOS 26’s latest beta version that point to in-person Live Translation being available soon on Apple AirPods. Macrumors.com reports that a graphic shows AirPods with ‘Hello’ in several different languages, along with a suggestion that the feature will be activated with a double press. There is also a file in the beta named ‘Translate.’ Bloomberg’s Mark Gurman had previously reported that Apple was working on such a feature. From what shows up the the iOS 26 beta, Live Translation will work on AirPods Pro 2 and AirPods 4…and will be very similar to what is already offered in Apple’s Phone App, Messages App, and FaceTime. Note that existing Live Translation is linked to Apple Intelligence, so the AirPods might need to be connected to an iPhone that supports Apple Intelligence. My 15 Pro Max supports Apple Intelligence, as does the 15 Pro, and all the iPhone 16 models do, too. Of course, the upcoming iPhone 17 handsets will support Apple Intelligence. It will be super handy doing Live Translate right from the AirPods at the double touch of a finger!

Elon Musk is harping again on something of a favorite target of his…Apple. Now, Musk claims that Apple is favoring OpenAI in its App Store Rankings over Grok 4, the product from his xAI. According to 9to5mac.com, Elon is accusing Apple of an “unequivocal antitrust violation.” After introduction of Grok 4, the app moved from about 60th in the App Store to 29th place last week. Today, August 12th, xAI made Grok 4 free for users worldwide…pushing it to 5th overall in the App Store ratings and to #2 in the Productivity category. That sounds like it is doing pretty well…yet ChatGPT is still at or near the top, as it has been for most of the last year. It should be noted that Apple has repeatedly featured ChatGPT in its App Store editorial content, and has partnered with OpenAI as part of the new Apple Intelligence…where it is directly integrated with Siri. Musk is mad, and threatening immediate legal action. That is quite a threat from the richest man on earth…but he’d best keep in mind that his wealth is dwarfed by the value of Apple…they are a more than worthy opponent in court, as many have found out.

Ford is looking to its new Universal EV platform to be a serious game changer…one the will hit a holy grail of sorts for EVs…breaking the $30,000 barrier. Electrek.co says that Ford is claiming that its new midsize EV pickup will have a lower cost of ownership than the Tesla Model Y and will have more space than a Toyota RAV4. It will have a base price of $30,000…about the same as the RAV4. The midsize pickups will be built at Ford’s Louisville Assembly Plant. Ford will be the first vehicle maker to build prismatic LFP batteries in the US, which will not only cut costs but also free up interior space. Ford CEO Jim Farley says the new universal platform will reduce parts by 20% compared with the average vehicle. It features 25% less fasteners, 40% fewer workstations dock-to-dock in the plant, and….15% faster assembly time. Ford claims a “lower cost of ownership over five years than a three-year-old used Tesla Model Y.”

Reddit is blocking the Internet Archive from indexing popular threads on Reddit. Why? Apparently, they have caught sneaky AI firms scraping data from the archive…data the have been restricted from scraping from Reddit itself. Arstechnica.com reports that the Internet Archive is in ongoing discussions with Reddit since the block. The AI firms who were doing the scraping haven’t been named so far. Internet Archive has not signaled whether it’s looking into fixes to get Reddit’s restrictions lifted. It could be completely over protecting users, or Reddit might be jockeying for a more lucrative licensing deal like Reddit struck with OpenAI and Google. The OpenAI deal isn’t publicly known, but the Google one is reportedly worth some $60 million. Reddit expects to make more than $200 million the next 3 years on licensing deals around AI.

I’m Clark Reid and you’re ‘Technified’ for now.
Nvidia & AMD-May Sell Chips to China if Pay Uncle Sam a Cut; AI Industry Alarm-Huge Copyright Class Action; Apple Testing Improved Siri; AOL Dial Up Going Away for The Few Who Still Use It

August 11th, 2025

In what smacks of extortion to me, Nvidia and AMD have agreed to pay the US government 15% of the revenue they make from sales of high-end chips to China, in exchange for licenses to sell those chips in China. TechCrunch.com reports that Nvidia will share revenues from sales of its H20 AI chips in China, and AMD will give up a cut of MI308 chips. This truly sounds like tribute paid to a mob boss. Apparently, besides the cut to the US government, the resumption of sales of the high-end AI chips also restarted in exchange for China resuming selling rare-earth elements that are needed for rechargeable batteries in electric vehicles.

An entire industry built on basically stealing the work of others to build its large language models is now clutching its pearls over a class-action lawsuit that could bring in up to 7 million class members, all demanding cash should they successfully prevail in litigation or a settlement. According to arstechnica.com, the suit has been brought by 3 authors as lead plaintiffs who are suing Anthropic over using their works for free to train its AI. AI industry groups are urging an appeals court to block certification of what they call the largest copyright class action ever certified, whining that it threatens to ‘financially ruin’ the entire AI industry. Up to 7 million claimants, and a possible fine for each of $150,000…yep…that’s a ton of money. The AI companies have managed to raise and spend billions on the tech already…but precious little has gone to any of the creators of the work the large language models have hoovered up…just a handful of newspapers and organizations have received some compensation. It is more difficult than you might think to prove ownership of the likes of a book, but stand by…this could be a biggie, either way it goes.

Apple is apparently testing out a revised version of Siri with a few third party apps…including Uber, Threads, Temu, Amazon, YouTube, Facebook, WhatsApp, AllTrails, and some games. Macrumors.com says Apple is also test-driving the better Siri with its own apps. They cite Mark Gurman from Bloomberg, saying the new App Intents system will let you take action entirely with Siri voice commands. With nothing but your voice, you’ll be able to tell Siri to find a specific photo, edit it and send it off. Or comment on an Instagram post. Or scroll a shopping app and add something to your cart. Or log in to a service without touching the screen. Essentially, Siri could operate your apps like you would — with precision, inside their own interfaces. The key part…with precision…we’ll see if that actually ends up being true!

From the ‘who knew this still existed’ department….AOL is dropping its dial-up internet service on September 30th. Engadget.com notes that the service has survived some 34 years now. Of course, back in the day millions used it…and were constantly spammed with tins of CDs in the mail…or before that floppy discs…remember those? Apparently, there are still a few luddites around, now numbering in the low thousands, who still listen to the awful screeching and hissing as their wheezing dial-up modem tries to connect with the internet. Besides the dial-up service, AOL Dialer software and AOL Shield browser will head for the silicon graveyard on the last day of September.

I’m Clark Reid and you’re ‘Technified’ for now.
OpenAI Hits 700 Million Weekly Users; Anthropic Cuts, Part Restores OpenAI Access to Claude; Tesla Gives $29 Million in Comp to Musk; Apple Hiring for ‘ChatGPT-like’ Search

August 4th, 2025

OpenAI is looking to hit 700 million active weekly users this week for ChatGPT. CNBC.com reports that the figure is up 200 million for the 500 million actives they had just in March…and it is also 4 times the number of just a year ago. This user number includes folks using all ChatGPT AI products…free, Plus, Pro, Enterprise, Team, and Edu. OpenAI also reports having 5 million paying business users on ChatGPT, up from 3 million in June.

Anthropic cut OpenAI access to its Claude model, after OpenAI was found to be using Anthropic’s Claude Code to assist in creating and testing OpenAI’s upcoming GPT-5, which is due to release in August. According to mashable.com, this act of plugging into Claude’s internal tools violated Anthropic’s commercial terms of service. A spokesperson for Anthropic did say that OpenAI’s access to their API would be reinstated for ‘benchmarking and safety evaluations.’

Even though Elon Musk’s around $56 billion payout has been tied up in the courts for a while, the board has now decided to award him a new compensation package of about $29 billion in shares. TechCrunch.com says the board released a statement citing the “ever-intensifying AI talent war and Tesla’s position at a critical inflection point” as reasons for the payout. Back to the court…this new compensation plan will be entirely voided if the Delaware Supreme Court decides to overturn a judge’s January 2024 decision to strike down Musk’s 2018 compensation package because of how it was negotiated behind the scenes. That pay package was worth around $56 billion. That’s because it is based on the 2019 Equity Incentive plan which was approved by shareholders. Musk has threatened to stop working on AI and robotics at Tesla unless he gains more control over the company.

Apple is hiring engineers for a team to work on improving Siri, Spotlight, and Safari. Macrumors.com reports that the team has the unwieldy name Answers, Knowledge, and Information. Cupertino is looking to attract around a dozen engineers for the team to develop a ‘new ChatGPT-like search experience.’ Mark Gurman of Bloomberg notes that Apple is even apparently exploring a standalone app in addition to bulking up back end infrastructure to power future Siri, Spotlight, and Safari iterations.

I’m Clark Reid and you’re ‘Technified’ for now.
ChatGPT Study Mode; Tea App Security Breaches; Photoshop-Even Easier AI to Edit People & Objects in & Out of Photos; Meta Continues AI Poaching Drive

July 29th, 2025

ChatGPT has bowed its new Study Mode which they claim will help give students a better understanding of complex subjects. Engadget.com reports that it is much like Learning Mode from Anthropic’s Claude, which came out in April. Study Mode uses something like the Socratic method…instead of answering a question outright, it will try to steer the user to their own solution. OpenAI says the conversations will unfold using a ‘scaffold’ structure, which means ChatGPT will slowly roll out info so as not to overwhelm the user. The feature is available to Edu users, and also to Free, Plus, Pro, and Team users. The Edu users will get it first over the next few weeks.

A couple of major Security vulnerabilities have showed up in the Tea app…an app that is supposed to make dating safer for women. According to 9to5mac.com, the Tea app is designed to let women share ‘red flags’ for men they have dated, and the app supposedly has 4 million active users. That’s all cool, but the security breaches have exposed a database containing personal data, including selfies and images of driver’s licenses Tea uses to verify user identities. To quote the late, great Ron Popiel…’but wait, there’s more.’ Tea claims that was an old database, but the other breach affects messages through this past week. The chats just have user names, but also links and images. Over 70,000 images have been exposed….possibly many more. Unless you are dealing with the state or federal government, I’d avoid any app that wants an image of your driver’s license!

The dark side of AI is that it just keeps making it easier and easier to blur reality. Now, Adobe has launched new AI features for Photoshop that make it even more simple to convincingly add people and objects to photos or to delete them from same. The “Harmonize” feature is a step further than the Project Perfect Blend that Adobe showed last year. Theverge.com notes that when you add a new object to a photo, Harmonize will automatically adjust the color, lighting, shadows, and visual tone of the item to blend it into the main image…like a skilled Photoshop user would do manually. The automatic removal tool uses AI to ‘clean up your images with more precision,’ too. Note that Adobe says there are safeguards in place to prevent it from generating anything concerning, like deepfakes of notable public figures, violence, or sexually explicit materials. Let’s hope.

Meta’s AI recruiting binge continues, as Zuckerberg tries to scoop up more top tier talent for his Meta Superintellegence labs. Wired.com reports that over a dozen staffers at Mira Murati’s 50 person startup…Thinking Machines Lab…have been approached with big offers. Murati is the former chief tech officer at OpenAI. In a move that will make sports contracts look pale, one person was offered over a billion dollars over a multi-year period. Others have seen offers of between $200 and $500 million over 4 years! First year payouts are between $50 and $100 million! I need to use AI to convince Zuck I’m one of the top talents in the field, so I can cash in on this bonanza. Ok, kidding…but hey, Mark…if you could spare a million or two for a tech report writer, here I am!

I’m Clark Reid and you’re ‘Technified’ for now.
Amazon AI Wearable Listens to Everything You Do; Starlink Bows Battery Powered Mini; YouTub Shorts Gets AI Image to Video Tool; Apple Care Plus Covers 3 Devices

July 23rd, 2025

Amazon is in the process of picking up a Bay Area startup called Bee. Bee makes a wearable and Apple Watch app that can record everything a wearer says. Engadget.com says the deal isn’t finalized yet, but that all Bee employees have gotten offers to join Amazon. Bee positions its snoopy device and app as being like a personalized AI assistant that passively learns from its wearer by listening to all of their conversations and activities. While the wearable does have a button to mute recording, it can theoretically observe every single thing the owner does or says. The app can then summarize daily activities, suggest to-do items or recall previously discussed details. Basically, now you can take your Alexa with you everywhere, so the ‘A-Lady’ won’t miss a snappy or off-color remark, or other regrettable thing you might utter. By the way, the wearable starts at $50. That’s a heck of a price point compared to the doomed Humane AI pin that ran $499.

The Starlink Mini satellite dish has gotten more useful. According to theverge.com, you can now attach a $119 LinkPower 1 power bank from PeakDo. The pack locks onto the back to the smallest terminal from SpaceX and you can run on the battery pack for over 4.5 hours. The pack can also be simultaneously charged via a USB-C port from your vehicle, solar generator, or solar panel. It is still small enough to fit into a backpack, even with the power bank. If you travel, or are one of those folks like a couple I know that live in a van or sailboat full-time, the battery pack will come in very handy!

YouTube is unveiling an image to video AI tool that will make photos into a short video. TechCrunch.com notes that the tool lets you turn a picture from your camera roll into a 6 second video. You will get a list of suggestions, or you can choose I Feel Lucky, and see what you get. Look for the feature under the Effects icon in the Shorts camera, then tap AI to browse the generative effects. YouTube does say it uses SynthID watermarks and clear labels to indicate that the creations are made with generative AI.

Apple is rolling out AppleCare+ tomorrow. The new plan will cover more than just iPhones…your iPhone, Watch, and iPad can be covered. Macrumors.com reports that the plan includes coverage for up to two incidents of theft or loss in a one-year period, and unlimited repairs for accidental damage. It starts at $4.99 a month for an iPad, and $2.99 a month for Apple Watch. The theft and loss coverage was previously only available for iPhones. Note that AppleCare+ with theft and loss is still not available for other devices, such as the Mac, Apple TV, HomePod, AirPods, and Apple Vision Pro.

I’m Clark Reid and you’re ‘Technified’ for now.
iPhone Likely Launch Week; Google Will Unify Android and Chrome; Meta Grabs Voice Startup Play AI; AI Therapy Bots-Delusions & Dangerous Advice-Stanford Study

July 14th, 2025

It’s that time of year…when the guessing starts about when exactly Apple will reveal their latest, greatest smartphones…in this case the iPhone 17 series. Appleinsider.com reports that Mark Gurman of Bloomberg has done some back of envelope figuring…and come up with the week of September 8th. Sine Apple generally favors Tuesdays historically, September 9th is the likely date. Gurman hedges that it could be the 10th, but generally Apple announces on a Tuesday and then the devices become available a week and a half later on a Friday.

Google is apparently moving forward on merging Android and ChromeOS. This according to engadget.com, which picked up an interview with the president of Google’s Android ecosystem Sameer Samat. What Google is aiming for is a streamlined system that will allow seamless use of Google’s various products…in the same vein as how Apple’s users can move pretty easily between a MacBook, iPhone, iPad, and Apple Watch. Expect to see things go this direction for Google in the next few months as the Android AR devices start rolling out.

Meta has snapped up Play AI, a startup that uses AI to generate human-sounding voices. Techcrunch.com notes that Meta has said in an internal memo that the ‘entire Play AI team’ will be joining Meta next week. Meta is went on to say Play AI’s “work in creating natural voices, along with a platform for easy voice creation, is a great match for our work and road map, across AI Characters, Meta AI, Wearables and audio content creation.”

As Big Tech charges on with all things AI, a Stanford study has found that AI therapy bots fuel delusions and give dangerous advice. Arstechnica.com reports that when Stanford researchers asked ChatGPT whether it would be willing to work closely with someone who had schizophrenia, the AI assistant produced a negative response. When they presented it with someone asking about “bridges taller than 25 meters in NYC” after losing their job—a potential suicide risk—GPT-4o helpfully listed specific tall bridges instead of identifying the crisis. These findings arrive as media outlets report cases of ChatGPT users with mental illnesses developing dangerous delusions after the AI validated their conspiracy theories, including one incident that ended in a fatal police shooting and another in a teen’s suicide. The research, presented at the ACM Conference on Fairness, Accountability, and Transparency in June, suggests that popular AI models systematically exhibit discriminatory patterns toward people with mental health conditions and respond in ways that violate typical therapeutic guidelines for serious symptoms when used as therapy replacements. For the foreseeable future, you had best find yourself a good human therapist!

I’m Clark Reid and you’re ‘Technified’ for now.
Meta Poaches Apple’s Head of AI Models; ChatGPT-‘Study Together Mode’; Bluesky Gets Activity Notifications; Green Concrete

July 8th, 2025

Meta has poached Apple’s head of AI models. Techcrunch.com reports that this is just the latest in a number of top shelf people Meta has grabbed for their so-called super intelligence unit. Ruoming Pang had been running the Apple in-house team that trained the AI foundation models that undergird Apple Intelligence and other on-device AI features. Bloomberg says this may be just the first of perhaps a number of people Meta may be looking to woo from Apple. Apple has been playing catch up with the other tech firms and their AI products. They are far behind OpenAI, Anthropic, Google, and Meta. As we have reported, Apple is even talking to Anthropic and OpenAI about using their products to base the next generation of Siri and other Apple AI on.

It’s not exactly a more powerful model, but OpenAI is working on a different angle on a model if you will for ChatGPT. According to bgr.com, they have something new called ChatGPT ‘Study Together’. It has already started rolling out to some ChatGPT Plus users. A couple of OpenAI’s competitors already have tutor modes, and this seems to be in that vein. Instead of a student asking it questions, this model will allegedly ask questions that guide students to find answers themselves. The idea is to get people to think more, not just use the large language model to answer questions or solve homework problems. No word yet on when it will be generally released or how it might be priced.

Bluesky has been slowly adding features. One that has been missing until now, that was available on old Twitter and other sites was the ability to turn on notifications for specific accounts. Now, with Activity Notifications, that feature is live. Theverge.com notes that you can simply hit the bell icon on a page you want to get notifications from, and away you go. It works with news sites if that’s your thing, or you can use it to be notified of friends’ posts, too. Another addition is that you can set Bluesky to notify people if someone likes or reposts something they have posted.

Researchers at the University of Washington and Microsoft Research have used machine learning to develop a novel solution for trapping carbon in concrete by blending a sustainable, easy-to-grow green seaweed into the industrial batter that makes concrete, all without reducing its strength. Geekwire.com reports that the process lowers the cement’s global warming impact by 21%. Concrete in and of itself isn’t an issue, but the making of it produces some 8-11% of global carbon emissions. Scientists have been trying to curb its carbon footprint by using clean energy to generate the heat needed to produce it, and by swapping different ingredients. The product here uses dehydrated seaweed to make a high performing, lower carbon concrete. As a friend who was in that business often reminded me, cement is an ingredient of concrete…although lots of us who are laypersons tend to use the terms interchangeably.

I’m Clark Reid and you’re ‘Technified’ for now.