Biz & IT / Informed technology
-
97% of CrowdStrike systems are back online; Microsoft suggests Windows changes
Kernel access gives security software a lot of power, but not without problems.
-
At the Olympics, AI is watching you
New system foreshadows a future where there are too many CCTV cameras for humans to physically watch.
-
Hang out with Ars in San Jose and DC this fall for two infrastructure events
Join us as we talk about the next few years in AI & storage, and what to watch for.
-
Google claims math breakthrough with proof-solving AI models
AlphaProof and AlphaGeometry 2 solve problems, with caveats on time and human assistance.
-
OpenAI hits Google where it hurts with new SearchGPT prototype
New tool may solve a web-search problem partially caused by AI-generated junk online.
-
Chrome will now prompt some users to send passwords for suspicious files
Google says passwords and files will be deleted shortly after they are deep-scanned.
-
Secure Boot is completely broken on 200+ models from 5 big device makers
Keys were labeled "DO NOT TRUST." Nearly 500 device models use them anyway.
-
We made a cat drink a beer with Runway’s AI video generator, and it sprouted hands
Gen-3 Alpha produces wild and whimsical results. Here's what it cooked up for us.
-
CrowdStrike blames testing bugs for security update that took down 8.5M Windows PCs
Company says it's improving testing processes to avoid a repeat.
-
Elon Musk claims he is training “the world’s most powerful AI by every metric”
One snag: xAI might not have the electrical power contracts to do it.
-
How Russia-linked malware cut heat to 600 Ukrainian buildings in deep winter
The code was used to sabotage a heating utility in Lviv at the coldest point in the year.
-
The first GPT-4-class AI model anyone can download has arrived: Llama 405B
"Open source AI is the path forward," says Mark Zuckerberg, misusing the term.
Paul Sutter walks us through the future of climate change—and things aren’t great
This episode of Edge of Knowledge focuses on our rapidly transforming world.
-
Microsoft says 8.5M systems hit by CrowdStrike BSOD, releases USB recovery tool
When reboots don't work, bootable USB sticks may help ease fixes for some PCs.
-
Astronomers discover technique to spot AI fakes using galaxy-measurement tools
Researchers use technique to quantify eyeball reflections that often reveal deepfake images.
-
CrowdStrike fixes start at “reboot up to 15 times” and get more complex from there
Admins can also restore backups or manually delete CrowdStrike's buggy driver.
-
Major outages at CrowdStrike, Microsoft leave the world with BSODs and confusion
Nobody's sure who's at fault for each outage: Microsoft, CrowdStrike, or both.
-
The next Nvidia driver makes even more GPUs “open,” in a specific, quirky way
You can't see inside the firmware, but more open code can translate it for you.
-
OpenAI launches GPT-4o mini, which will replace GPT-3.5 in ChatGPT
Lower-cost AI language model will be free for ChatGPT users.
-
Vulnerability in Cisco Smart Software Manager lets attackers change any user password
Yep, passwords for administrators can be changed, too.
-
Trump allies want to “Make America First in AI” with sweeping executive order
After repealing Biden's AI order, draft would create "Manhattan Projects" for military AI.
-
Rite Aid says breach exposes sensitive details of 2.2 million customers
Stolen data includes customer names, addresses, birth dates, and driver's license numbers.
-
Former OpenAI researcher’s new company will teach you how to build an LLM
Karpathy's Eureka Labs will pair human-made curriculum with an AI-powered assistant.
-
Here’s how carefully concealed backdoor in fake AWS files escaped mainstream notice
Files available on the open source NPM repository underscore a growing sophistication.
-
Microsoft CTO Kevin Scott thinks LLM “scaling laws” will hold despite criticism
Will LLMs keep improving if we throw more compute at them? OpenAI dealmaker thinks so.
-
Google makes it easier for users to switch on advanced account protection
The strict requirement for two physical keys is now eased when passkeys are used.
-
OpenAI reportedly nears breakthrough with “reasoning” AI, reveals progress framework
Five-level AI classification system probably best seen as a marketing exercise.
-
First “Miss AI” contest sparks ire for pushing unrealistic beauty standards
Influencer platform's controversial contest awarded prizes to three nonexistent people.
-
Exim vulnerability affecting 1.5M servers lets attackers attach malicious files
Based on past attacks, it wouldn’t be surprising to see active targeting this time, too.
-
Intuit’s AI gamble: Mass layoff of 1,800 paired with hiring spree
Intuit CEO: "Companies that aren’t prepared to take advantage of [AI] will fall behind."
-
Threat actors exploited Windows 0-day for more than a year before Microsoft fixed it
The goal of the exploits was to open Explorer and trick targets into running malicious code.
-
In bid to loosen Nvidia’s grip on AI, AMD to buy Finnish startup for $665M
The acquisition is the largest of its kind in Europe in a decade.
-
OpenAI board shake-up: Microsoft out, Apple backs away amid AI partnership scrutiny
Microsoft gives up non-voting observer board role; Apple rethinks a planned similar position.
-
How disinformation from a Russian AI spam farm ended up on top of Google search results
A fake article about Volodymyr Zelensky’s wife buying a Bugatti with US aid was promoted by bots.
-
New Blast-RADIUS attack breaks 30-year-old protocol used in networks everywhere
Ubiquitous RADIUS scheme uses homegrown authentication based on MD5. Yup, you heard right.
-
The president ordered a board to probe a massive Russian cyberattack. It never did.
The Cyber Safety Review Board missed an opportunity to prevent future attacks, experts say.
-
384,000 sites pull code from sketchy code library recently bought by Chinese firm
Many website admins, it seems, have yet to get memo to remove Polyfill[.]io links.