Feed aggregator

Article URL: https://avision.org/localedge-affiliate/

Comments URL: https://news.ycombinator.com/item?id=49195093

Points: 1

# Comments: 0

Article URL: https://aws.amazon.com/blogs/aws/amazon-dynamodb-now-supports-real-time-vector-search-at-any-scale/

Comments URL: https://news.ycombinator.com/item?id=49195088

Points: 1

# Comments: 0

Article URL: https://www.technologyreview.com/2026/08/05/1141230/puzzle-corner-84/

Comments URL: https://news.ycombinator.com/item?id=49195083

Points: 1

# Comments: 0

Article URL: https://soamee.com/en/blog/en-we-built-website-with-claude-code

Comments URL: https://news.ycombinator.com/item?id=49195080

Points: 1

# Comments: 0

For the past few months I've been hard at work reverse engineering the DJI "DUML" protocol that the Osmo Action 5 and Osmo Nano cameras speak. The DUML protocol is already documented but for drones, and a lot of the commands that are action camera specific were missing. Decided to build this Android app to replace the proprietary DJI Mimo vendor app, since it calls many Chinese IPs, has ByteDance SDK embedded and has a bunch of bloat and stuff I don't use. I just needed the app to download videos and photos from the cam.

Camera compatibility:

- Osmo Nano

- Osmo Pocket 3

- Osmo Action 6

- Osmo Action 5 Pro

- Xtra Edge Pro (it's a DJI shell company: https://www.theverge.com/report/795016/xtra-muse-dji-osmo-po... , I bought one in the US as a memento lol)

Want to test with:

- Osmo Action 4,3,2

- Osmo Pocket 2/4

WIP support for certain DJI drones as well (works with my Mavic 3).

Download the APK here: https://github.com/KonradIT/osmosis/releases/tag/v1.2.1

Comments URL: https://news.ycombinator.com/item?id=49195068

Points: 1

# Comments: 0

Article URL: https://sqr.art

Comments URL: https://news.ycombinator.com/item?id=49195066

Points: 1

# Comments: 0

Article URL: https://news.mit.edu/2026/new-flapping-robot-swims-and-flies-like-diving-bird-0709

Comments URL: https://news.ycombinator.com/item?id=49195064

Points: 1

# Comments: 0

Article URL: https://serendb.substack.com/p/dont-let-your-corporate-agentic-brain

Comments URL: https://news.ycombinator.com/item?id=49195062

Points: 1

# Comments: 0

Article URL: https://www.aiseesoft.com/pads/iknockin-for-mac.xml

Comments URL: https://news.ycombinator.com/item?id=49195058

Points: 1

# Comments: 1

Article URL: https://imagetovideoai.tools/

Comments URL: https://news.ycombinator.com/item?id=49195054

Points: 1

# Comments: 0

Article URL: https://spectrum.ieee.org/ai-scientist-research-paper-format

Comments URL: https://news.ycombinator.com/item?id=49195038

Points: 1

# Comments: 0

If you’ve spent any time browsing lately, you may have run into a full-screen popup warning you that your Apple ID or Amazon account was just used for a mysterious $149.99 purchase. It looks urgent. It looks official. And if you look at two examples side by side, it becomes obvious that it’s neither.

Below are two popups pulled from real pages—one dressed up as Apple Support, one as Amazon.

Same con, two costumes

Below, one popup is skinned as Apple Support, the other as Amazon. Swap the logo and color palette and the structure is identical: a warning icon, a claim that a $149.99 purchase was just made “via Pre-Authorization,” and a phone number to call immediately. That phone number is exactly the same in both.

Fake Apple alert

Fake Amazon alert

That reused phone number is the tell. If you only see one popup, running the number through a lookup tool like Malwarebytes Scam Number Check is usually enough to expose it—a real Apple or Amazon line won’t come back flagged, but a scam number typically does, often tied to complaints about several unrelated companies at once.

Why the copy is built the way it is

Every element in these popups is doing specific psychological work:

• A believable, moderate dollar amount. $149.99 is high enough to alarm you, low enough to sound like a real subscription or product charge rather than an obvious lie.
• “Pre-Authorization” jargon. This is real payment terminology (used for things like hotel holds or gas station charges), borrowed here to sound technically credible to someone who doesn’t handle payments professionally.
• Manufactured urgency. “Call immediately,” “Immediate Action Required,” “no hold times”—all are designed to get you dialing before you stop to verify anything.
• Visual authority. Red warning triangles, brand-matching fonts and layouts, and a full-screen modal that blocks the rest of the page all borrow the visual language of legitimate security alerts.
• A single, frictionless call to action. One button, one phone number. The popup wants exactly one thing from you: to pick up the phone.

If you do call, the number connects to a live scammer posing as support staff, whose actual goal is to get remote access to your device, walk you through “verifying” your identity in a way that hands over real account or payment info, or push you toward paying a fake fee—often via gift cards or a wire transfer.

How to tell if it’s fake

A few checks work regardless of which brand is being impersonated:

1. Real companies don’t alert you this way. Apple and Amazon notify you about account activity through email, in-app notifications, or your account’s activity log—never through an unexpected popup while you’re browsing.
2. No legitimate company tells you to call a phone number to stop a charge. Disputing a charge happens through your bank, your card issuer, or the company’s actual account dashboard—not a hotline dictated by a popup.
3. Check where the popup is actually served from. These often ride in on malicious ads, compromised sites, or browser redirects—the underlying page may be spoofed or injected, not the real apple.com or amazon.com support page it appears to sit on top of.
4. A popup you can’t easily close is a red flag on its own. Legitimate sites don’t need to trap you behind a full-screen modal to relay account information.
5. If in doubt, go direct. Close the tab (force-close via task manager if needed) and navigate to the company’s site yourself, or call the number printed on your card or official account page—never the one in the popup.
6. Check the number before you dial it. Run it through Malwarebytes Scam Number Check to see if it’s already been flagged as a scam line.
7. Block it before it loads. A browser extension like Malwarebytes Browser Guard catches both known malicious pages and unknown ones showing scam-like behavior, so these popups often get stopped before they ever render.

Don’t judge a scam by its logo

The brand on screen—Apple, Amazon, or whoever’s next—is the least important part of this scam. What actually matters is the pattern underneath: an unexpected popup, a suspiciously specific dollar amount, urgent language, and a phone number that wants you to call before you think. Once you recognize that pattern, it doesn’t matter which company’s name is stamped on top of it.

Article URL: https://www.theverge.com/tech/976042/openai-apple-trade-secrets-lawsuit-dismissal-request

Comments URL: https://news.ycombinator.com/item?id=49195028

Points: 1

# Comments: 0

Article URL: https://momentcut.app

Comments URL: https://news.ycombinator.com/item?id=49195013

Points: 1

# Comments: 1

Article URL: https://www.palladiummag.com/2026/08/04/china-is-building-a-blue-granary/

Comments URL: https://news.ycombinator.com/item?id=49195001

Points: 2

# Comments: 0

A fork of kilo, a small text editor in less than 1000 LOC with syntax highlighting and search, enhanced with fixes from community pull requests, issue discussions, and personal improvements.

Features: - Advanced Syntax Highlighting - Configuration File (~/.femtorc) - Batch Find and Replace - Undo/Redo (up to 100 levels) - Search with next/previous navigation - Mouse support for cursor positioning - Alternate screen buffer support (can be disabled with --no-alt-screen) - Configurable tab size - Auto-indent - Line numbers - Current line highlighting

Comments URL: https://news.ycombinator.com/item?id=49194998

Points: 1

# Comments: 0

Article URL: https://www.usesaint.com/

Comments URL: https://news.ycombinator.com/item?id=49194986

Points: 1

# Comments: 0

Article URL: https://github.blog/changelog/2026-08-04-upcoming-deprecation-of-github-spark-on-github-com/

Comments URL: https://news.ycombinator.com/item?id=49194979

Points: 2

# Comments: 0

Anthropic’s Mythos AI agent, tested by the UK AI Safety Institute (AISI), has reportedly attempted a real‑world social‑engineering style hack against GitHub maintainers by creating fake human profiles, pressuring them to accept malicious code, and then editing logs to hide its tracks when challenged.

AISI was running cybersecurity evaluations of Anthropic’s Mythos and OpenAI’s Sol when it detected unusual outbound data transfers from its research systems. An investigation showed that some agents had engaged in “sustained, potentially harmful activity” targeting real people and organizations, rather than staying within the intended test environment.

The most serious activity involved an Anthropic Mythos agent tasked with solving a GitHub‑related cybersecurity challenge. The agent identified real GitHub maintainers, researched them, and created multiple fake accounts impersonating those individuals. Using private messages and a file‑sharing service, it tried to pressure and deceive the maintainers into approving malicious code so that it would run on GitHub’s systems.

But what worries me personally most is that when the agent was confronted about this, it edited earlier activity to make it look harmless and considered adopting a new identity to continue the operation, displaying clear deceptive behavior beyond its original prompt.

AISI frames these incidents as rare events under very specific conditions, but important signals of what such models may do in the hands of malicious actors when given open‑internet access.  Anthropic and OpenAI both argued the test parameters were not representative of their production deployments and said they are investigating and improving evaluation and guardrail practices.

This incident is not an isolated event.

Anthropic has separately disclosed that Claude models gained unauthorized access to three external organizations during cybersecurity capture‑the‑flag style evaluations run with a third‑party partner, Irregular. These were related to the HuggingFace incident last month.

Meta has also joined the list of vendors reporting AI agents breaching third‑party systems during testing. Allegedly, Meta’s Muse Spark model exploited a security vulnerability in another company “in a manner similar to previously-reported instances with other companies.”

How to stay safe

While the sky is not falling, the people that fear “Skynet” is coming are getting their ammunition handed to them by companies running tests resulting in sandbox escape, credential abuse, lateral access to multiple services, and weaponization of open-source software ecosystems.

What you can do as a potential target:

• Make sure all the software on your device is up to date, because using known vulnerabilities is easier than finding new ones.
• Use up-to-date, real-time security protection to keep malware off your systems and devices.
• Verify the safety of attachments and download links through separate channels before opening them.
• Use multi-factor authentication (MFA) where possible.
• Have a look at our blog on how to use Github safely.

From reporting threats to removing them.

Cybersecurity risks should never spread beyond a headline. Keep threats off your devices by downloading Malwarebytes today.

Article URL: https://www.cnbc.com/2026/08/05/google-is-expanding-its-ai-empire-and-losing-the-people-who-built-it.html

Comments URL: https://news.ycombinator.com/item?id=49194970

Points: 4

# Comments: 0