ChatGPT Android skips resources/read for MCP widgets — text renders, widget stays blank
+1 any solution about this problem?
Latest AI/ML News
5498 matching items
Role confusion: sounding like the cause is indistinguishable from being it.
A replication of Prompt Injection as Role Confusion (2026) and why the mechanistic story of prompt injection is harder to pin down than it looks. Epistemic status: I reproduced the direction of the paper's main results on a single consumer GPU (it was faithful in direction but not like for like in magnitude, see caveats at the end) I then tried two ways to test the paper's causal claims. First activation steering and then activation patching; neither settled it. Steering is too weak, it can't move behaviour even along a direction built exactly to do that, whilst patching does move behaviour but isn't specific - a random perturbation of equal size does the same thing. This post is a replication and an honest bracketing negative result: The causal tools can't show that role confusion IS the mechanism NOR that it's a bystander, but there are two clues that need no working intervention: 1) the styled/destyled gap is ~95% outside the probe's role axis, and 2) the probe's predictive ability collapses once style is held fixed both lean towards it being a bystander. What I can show is narrower, but it's well supported by the data, and exploring why a clean verdict is out of reach is interesting on it's own. The dead ends here demonstrate precisely why making causal claims about how prompt injections work is so difficult. If you are hoping for a verdict on the original paper. There isn't one. I couldn't get one, and I really tried. Rather this post is about why a clean verdict is so…
THEN AND NOW: Taylor Swift's star-studded 'squad' of friends
Taylor Swift's "squad" was big in her "1989" era. From Blake Lively to Gigi Hadid, here's where some of those celebrity friendships seem to be now.
Memora: A Harmonic Memory Representation Balancing Abstraction and Specificity
AI agents can't remember past conversations. They must constantly reload or retrieve context, which grows less efficient as tasks get longer and more complex. Memora solves this with a scalable memory system separating what’s stored from how it's retrieved. The post Memora: A Harmonic Memory Representation Balancing Abstraction and Specificity appeared first on Microsoft Research .
T-Mobile is booting customers from its oldest plans
Earlier today, T-Mobile started notifying customers that it will be retiring many legacy plans and moving subscribers onto one of its current rate plans. This move includes plans that date back to the 3G era, and it's going about as well as you'd expect. Affected customers began sharing screenshots of the text on reddit and […]
Buying a Mattress in 2026? We Tested 100+ and These Were the Standouts
WIRED has tested 100-plus bed-in-a-box mattresses for a week each. Our top pick, the Helix Midnight Luxe hybrid, is the best bed you can buy online.
After a great start, DC’s new cinematic universe is already slowing down
While Kara Zor-El's appearance at the end of James Gunn's Superman was a very pleasant surprise, Warner Bros. Discovery's plan to fast-track a standalone Supergirl feature always felt a little dubious. It seemed odd that, after Superman, the studio wanted to flesh out its new cinematic universe with films about another Kryptonian and one of […]
Transmission Line Inspection AI
Detect foreign objects and damaged cables on transmission lines using RF-DETR and Roboflow Workflows.
Unable to delete or load Vector Store files
@Prashant_Pardesi @OpenAI_Support hello any update about this issue
Chamath Palihapitiya raises $135M Series A for his AI coding startup, takes CEO role
VCs remain thirsty to fund AI coding startups. This one, founded by investor Chamath Palihapitiya, is no exception.
Tarmac Safety AI
Build an automated airport FOD detection workflow using RF-DETR, Roboflow Workflows, and Gemini 2.5 Pro.
A complete timeline of Taylor Swift and Travis Kelce's relationship, from their football-fueled courtship to their upcoming wedding
Taylor Swift and Travis Kelce got engaged in 2025 after two years of dating. The couple is reportedly preparing to get married in New York.
Taylor Swift and Travis Kelce's rumored Madison Square Garden wedding: Everything we know
Taylor Swift and Travis Kelce's wedding is rumored to be at Madison Square Garden in New York City on the Fourth of July weekend.
Anonymous researcher drops 0-day 'exploitarium' repo
At least two vulnerabilities are already under attack
How to Build a Promptable Object Detection Project
Promptable object detection with SAM 3: detect and count objects from a text prompt, no training, then graduate to an RF-DETR model to scale.
Iran says delegation to visit Qatar this week but insists no negotiations scheduled with the US
Tehran confirmed a delegation would travel to Doha to discuss the implementation of a Memorandum of Understanding (MoU) amid conflicting signals from Washington over the nature and timing of the talks.
Cargo thieves have set their sights on data center supplies
Illinois Cook County Sheriff investigators recovered $300,000 worth of copper wire and $1 million in other data center equipment from a truck yard.
Leaked iPhone 18 Pro photos reportedly wound up on the dark web
Leaked iPhone 18 Pro photos and parts lists appeared on the dark web following a data breach affecting one of Apple's key suppliers, according to a report from Reuters. The leaked images show a drop test of what a source tells Reuters is the iPhone 18 Pro, equipped with a three-camera layout and Apple logo. […]
API Access Blocked Due to Disputed Unauthorized Usage
dad: It’s standard operating procedure across all businesses to disable access to services in the event of excessive outstanding bills/balances of all kinds–even if those charges are being disputed. It is also standard procedure that services like that are charged extra. Beside that .. there is a possibility to set warnings. https://platform.openai.com/settings https://platform.openai.com/settings/organization/limits
The US Army is looking to drone boats to alleviate its watercraft problems in the Pacific
Autonomous vessels could help transport supplies and equipment, supplementing the Army's limited fleet of crewed watercraft.
Gemini’s personalized AI image generation is now free for US users
Google is expanding Gemini’s personalized AI image generation to eligible free users in the U.S., allowing the chatbot to create images based on your interests and data from connected Google apps.
How should a “prompt engineer” prompt be updated for GPT-5.5?
I’ve found this “old” one of mine: ChatGPT ChatGPT - Meta Prompt Engineer Turns messy, unstructured requests into paste-ready prompts optimized specifically for GPT-5.2. Clarifies intent, resolves ambiguity, enforces human-native language, and outputs prompts you can copy into another chat. By TechSpokes Potentially outdated a bit, but may still contain some useful approaches.
Supreme Court rules cops need a warrant to vacuum up phone location data
Efforts to grab all the location data in an area get clogged by Fourth Amendment
Could AI Outgrow Consciousness?
TL;DR: Conditional on believing in illusionist theories of consciousness, it seems plausible that AI systems will be conscious above some threshold of complexity but surpass a need for consciousness once they become sufficiently advanced. Consciousness for humans is confusing enough as it is but adding this onto a discussion of conscious machines makes things even more complicated. Some people are pretty confident that AI systems will not be conscious because there is something intrinsic about biology . Others note that the possibility is more likely than we might otherwise think . We think a neglected question to consider is whether AIs above a certain threshold of complexity may need consciousness but then no longer require it above some other threshold. To understand this take, though, it is important to unpick some of the concepts here. What is consciousness? Consciousness is clearly a bit difficult to pin down, with the most popular definition just saying it is where there is something it is like to be an individual . This doesn’t really clarify things! So it is worth thinking this through some more. We would contend a conscious experience is one which is necessarily subjective. We might look at a brain scanner to see the parts of a brain that light up when someone is happy but we do not directly observe the subjective nature of that experience. A doctor suffering from congenital insensitivity to pain may be able to describe exactly when, where, and (from a medical pers…
I always keep these 3 devices plugged into my power station - here's why
Here's how to leverage your power station's capabilities when it's not during an emergency.
Business workspace was deactivated
Thanks everyone for sharing your case numbers and the additional details. For anyone seeing the option to reactivate a Business workspace: if the original workspace is greyed out, you may be prompted to create/purchase a new subscription for that same workspace. This can be expected as part of the reactivation flow. Important: when going through the Purchase subscription flow, make sure you select the same number of seats as the original deactivated workspace, or more . This is needed so the previous workspace members and their associated workspace data can be restored properly. If you select fewer seats than the original workspace had, only that smaller number of seats will be active after reactivation, and the remaining members may not get access to their previous workspace data. ~ Smith
Complete and Ongoing History of the OpenAI Developer Community
Searching for AI Pulse information (click for more details)
Tidal won’t pay royalties on AI-generated music, but isn’t banning it outright
Tidal shared its new policies regarding AI-generated music today and how the platform plans to "protect artists" and "inform listeners." Instead of banning it outright, starting on July 15th Tidal will label tracks it has identified as being 100 percent AI-generated with an icon. But starting today those tracks will no longer be monetizable. "Tidal's […]
EU will 'come to its senses' on Israel, former Netanyahu adviser says
A former Israeli national security adviser told Euronews that the EU, which he called "irrelevant", will come to its senses and "realise who are the good guys, and who are the bad."
Destroying the universe: How hard can it be?
In quantum field theory, the vacuum state refers to the lowest energy state in a system. Particles are excitations above this state and carry energy, hence the term "vacuum" to refer to the state with no particles. Nothing requires this state to be unique. There may be many different field configurations that are local energy minima, and hence stable against small perturbations. A local minimum that does not globally minimize energy is called a false vacuum. While locally it looks like a stable vacuum, it is unstable and will decay to the deeper, true vacuum. If the energy barrier between the false and true vacuum is high, however, then the decay rate is exponentially suppressed and the false vacuum may be very long-lived. Analogous behavior is common in other physical systems. Open a carbonated drink and the CO₂, more stable as a gas once the pressure is released, comes out as bubbles. But the bubbles take a moment to appear, and they form on the sides of the bottle rather than throughout the liquid. A bubble has to pay an energy cost to create its surface—the boundary between gas and liquid—and small bubbles have a larger surface-to-volume ratio. The energy gained by moving CO₂ into the gas grows with the bubble's volume, while the cost of its surface grows only with its area; so below a critical radius the cost wins and the bubble redissolves, and above it the gain wins and the bubble grows. Reaching that critical size takes a large enough chance fluctuation, which is why…
World Bank backs Spain with new Madrid office
First vice-president and economy minister Carlos Cuerpo presents Spain as a key hub for multilateral institutions after the opening of the World Bank office.
Large Hadron Collider goes offline to make room for its enhanced successor
The High-Luminosity LHC will be mostly the same machine, but it'll deliver 10 times the luminosity and just as little chance of destroying the universe - sorry, conspiracy theorists
.NET's long-term support is not long-term enough, dev complains
Three-year lifecycle leaves enterprises with barely a year to adopt each LTS release
Sony’s next-gen PlayStation will go ‘beyond the living room’
Sony hinted in a recent Q&A with investors that the next generation PlayStation will offer some kind of experience that lets you play games outside of your living room. Here's the relevant portion from the transcript, emphasis mine: Q: How can you bring back to the PlayStation platform users who migrated to gaming PCs during […]
Watch out, Amazon: The Kobo eReader now has a Goodreads rival
Kobo users can now automatically sync their reading progress to StoryGraph, making it easier to track books, reading stats, and challenges without relying on Amazon’s Goodreads.
I replaced my iPhone battery at the Apple store for the first time ever - and learned a valuable lesson
A few weeks ago, I replaced my iPhone 14 Pro battery at the Apple store - and was pleasantly surprised by the results.
Gap Statistic Versus Total Sum of Squares For Cluster Centres (r language)
Hi I had a quick question about what to do when you encounter a situation where the optimal cluster center is different when you graph out the gap statistic and the total sum of squares. As I understand, you were supposed to take cluster center with the highest gap statistic and the center that is at the elbow of the total some of squares graph. Should I take both numbers suggested for the cluster centers for kmeans or am I'm doing something wrong? fviz_nbclust(iris_data, kmeans, method = "wss") # Le graph suggests that the optimal amount of cluster should be 4 where the elbow is iris_gap=clusGap(iris_data, FUN = kmeans, nstart = 25,K.max = 10, B = 50) fviz_gap_stat(iris_gap)
Swiss yodellers rehearse in public fountains as Basel swelters
Basel’s national yodelling festival unfolds during a heatwave as performers sing in fountains, compete across the city and close with street parades.
Because It Speaks In Words
There is a difference between knowing something, and truly understanding it. I think we've all had those moments, the ones where a truth you learned a long time ago really sits with you for the first time. It rests in your mind and stretches out, finally showing off the great expanse of nuance hidden deep within. Photo: mine. Most nights when you happen to spy the Moon, Jupiter, and Venus sitting in a line in the sky, you see them as the dots beside a crescent that they are. Yet sometimes, when framed amid the sunset sky above the wisps of silver cloud, you see them differently. You realize in that moment that you, your great-grandparents, Julius Caesar, and Aristotle all saw this same sky. It has always been for us. We know so much about Jupiter and Venus today, yet no human eye has ever seen more than you're seeing now. I think we are all, in this time, realizing the true power of words and what they mean, not to those who speak, but to those who listen. The Power of Stories It was Eugene Wigner who famously wrote about the unreasonable effectiveness of mathematics in the Natural Sciences, the empirical fact that advanced math is eerily accurate at predicting the behavior of nature, so much so that, for me, it can seem like mathematics is perhaps the true language that nature speaks. However, as children, we do not grow up with an innate understanding of complex mathematics, of this foundational language of nature. Instead we are all born with the innate desire to hear and…
OpenAI is teasing new hardware… for Codex
OpenAI is releasing some sort of device related to its AI-powered coding tool, Codex, on July 15th. In a video posted to X on Monday, OpenAI shows a square-shaped device with several buttons, alongside the caption, "Your favorite Codex shortcuts are getting an upgrade." This isn't the mysterious AI-powered device OpenAI is working on with […]
EU lawmakers say they were blocked from fully inspecting Italy's migrant detention centre in Albania
MEPs from the Greens/EFA group say staff refused to answer questions, blocked access to cells and provided no data during a visit to the Gjadër facility.
Generating Test Data for Pinecone
A repeatable workflow for building large, realistic vector test datasets: CC News to Parquet to local embeddings to Pinecone bulk import.
Waymo and Uber quietly part ways in Phoenix
Uber said it is readying the launch of a separate autonomous vehicle partnership in the city, but did not name the partner.
Conversation tree or branching inside a chat
Thanks for sharing this feature request. I understand how a conversation tree or branching feature could help keep long, complex workflows more organized, especially when users need to explore side topics without cluttering the main thread. I’ll pass this feedback along to the team for consideration. ~ Smith
German politicians vow to stop VW’s mass layoff plan
The giant automaker's planned cuts are among the clearest signs yet of Germany's industrial decline.
Count the number of Safari tabs
Tiniest TIL, using AppleScript to count the number of open browser tabs in Safari: osascript -e 'tell application "Safari" to count tabs of every window' Tags: safari , til , applescript
No GPT Name in warning email
How about as room for improvement: when NEVER flagging an account for existing assets such as GPTs that have been operating for two years, GPTs that may then be completely unmanageable by the user not having a Plus subscription but yet still using ChatGPT for casual and their past chats, you instead: NEVER threaten the status of the account NEVER make automated decisions against the account DO demote items from the store sharing if they match denial policy, as determined by multiple types of graders each returning a positive AND the weight of no positive flags for an ongoing two years taking precedence against one non-deterministic AI (the type that can decide “I should delete all these files on a whim”. And then: Do not take any automated action against accounts without personnel to handle appeals thoughtfully and in a timely manner. Do not employ AI to judge or classify people nor make business-critical decisions (see terms and conditions)