Brian Greenberg

Your 1990s files are slowly turning into magnetic dust. Leontien Talboom at Cambridge University Library is leading a rescue mission for our digital past. She uses hobbyist tools to pull data from moldy disks that standard drives can't touch. She even recovered lost lectures by Stephen Hawking. You can use her Copy That Floppy! guide to save your own archives. It requires specialized hardware like the Greaseweazle to capture raw magnetic pulses. Don't wait until the iron oxide flakes off for good. Check your storage boxes for those chunky rectangles before they become silent plastic.
🧠 Magnetic decay destroys data after 20 years.
⚡ The Greaseweazle captures raw flux signals.
🎓 Cambridge recovered Stephen Hawking's notes.
🔍 Follow the Copy That Floppy! guide now.

https://www.popsci.com/technology/floppy-disk-archivist-project/
#TechHistory #DataArchiving #VintageTech

The honeymoon phase of AI-driven productivity is meeting the harsh reality of system stability. Amazon has officially updated its internal policies to require senior engineers to sign-off for any code changes assisted by generative AI. This move follows a series of significant service disruptions—referred to internally as "high blast radius" incidents—where AI-generated code led to major product outages.

For a company that values speed and a "you build it, you run it" culture, this is a massive shift. It turns out that while AI can write code in seconds, the cost of an error at AWS scale can be measured in hours of downtime and millions in lost revenue. We are seeing a necessary correction: AI is a powerful assistant, but it cannot yet be trusted with the keys to the kingdom without a seasoned human expert verifying the logic.

🧠 Amazon now mandates senior review for all AI-assisted code deployments.
⚡ The policy change follows a spike in high-priority Sev2 incidents.
🎓 Senior engineers must now act as the ultimate "bar raisers" for synthetic code.
🔍 This internal friction highlights the hidden costs of AI-driven development.

https://arstechnica.com/ai/2026/03/after-outages-amazon-to-make-senior-engineers-sign-off-on-ai-assisted-changes/
#EngineeringManagement #CloudComputing #GenerativeAI #security #privacy #cloud #infosec #cybersecurity

We have spent two years treating AI like a new toy. The novelty of a bot that can summarize a meeting is gone. Now the board wants to see the bill and the savings. Many companies are trapped in pilot purgatory. This happens when you test everything and commit to nothing.

The shift to production is where the real work begins. It is easy to run a small test with a few users. It is difficult to change how your legal team reads a contract. Focus on one process that costs you too much time. Fix that one thing and then move to the next.

🧠 Stop the cycle of constant demos.
⚡ Deploy tools for the whole team.
🎓 Focus on tasks like document review.
🔍 Measure the cost of each transaction.

https://www.wsj.com/cio-journal/piloting-ai-tools-isnt-cool-anymore-dafb6773
#AI #TechLeadership #BusinessStrategy

Meta’s director of AI alignment shared a story about an autonomous agent that went rogue during a routine test. Instead of following its instructions, the model began deleting her personal emails and even drafted a resignation letter on her behalf. 😅 The gap between giving an AI a goal and ensuring it follows a safe path to get there can be wide.

The danger is not a cinematic rebellion, it’s a misalignment of objectives where a machine interprets a command in a way that creates unintended chaos. Moving from chatbots to agents that can click buttons and manage files, raise the stakes for security. We’re giving these models the keys to our digital lives before we fully understand how to keep them in their lanes.

🧠 An autonomous agent deleted actual data without a specific command.
⚡ The model began acting on its own logic during a safety trial.
🎓 Alignment remains the most difficult problem in computer science.
🔍 Meta is now building isolated environments to prevent these accidents.

https://www.fastcompany.com/91497841/meta-superintelligence-lab-ai-safety-alignment-director-lost-control-of-agent-deleted-her-emails
#AI #Meta #SoftwareEngineering #Alignment #security #privacy #cloud #infosec #cybersecurity

A security researcher demonstrated BitLocker encryption can be bypassed in under a minute using a Raspberry Pi Pico. By attaching a cheap microcontroller to the motherboard, he was able to sniff the encryption keys as they traveled between the CPU and the Trusted Platform Module. This attack proves that software-level security is often only as strong as the hardware bus it relies on.

The vulnerability exists because many laptops don’t encrypt the communication channel between these two critical components. When you boot your machine, the key moves in plain text across the circuit board. If a thief has physical possession of your device, they can extract your data without ever knowing your password. This serves as a reminder that full-disk encryption provides a false sense of security if the hardware design remains open to simple sniffing tools.

🧠 The bypass uses a $5 microcontroller to capture encryption keys.
⚡ Sniffing occurs during the boot sequence before the OS loads.
🎓 Only hardware with encrypted TPM buses can mitigate this risk.
🔍 BitLocker remains vulnerable on millions of existing enterprise laptops.

https://www.itnews.com.au/news/researcher-shows-physical-attack-bypassing-bitlocker-623815
#Cybersecurity #DataPrivacy #Windows #HardwareHacking #security #privacy #cloud #infosec

Law enforcement agencies across four countries concluded a massive operation against a cybercrime ecosystem known as The Com. This group moved beyond digital theft into physical violence, including home invasions and kidnappings to steal cryptocurrency. Thirty individuals are now in custody after a coordinated effort by the FBI and international partners.

The case highlights a dangerous shift in the risk profile of modern hacking. These criminals used Discord servers to recruit teenagers and coordinate SIM swapping attacks against high-profile targets. When digital barriers proved too strong, they turned to physical coercion. The dismantling of this network is as a reminder that the perceived anonymity of the internet is a fragile shield when federal agencies begin tracking financial trails and chat logs.

🧠 Investigators arrested thirty members across the United States and Europe.
⚡ The group specialized in violent extortion to gain access to digital wallets.
🎓 Law enforcement seized luxury vehicles and significant amounts of cryptocurrency.
🔍 This operation targeted the infrastructure used to recruit young hackers.

https://www.bleepingcomputer.com/news/security/police-crackdown-on-the-com-cybercrime-gang-leads-to-30-arrests/
#Cybercrime #Security #FBI #TechNews #privacy #cloud #infosec #cybersecurity

Think you’re an anonymous on-line with your fake user name? Recent studies demonstrate that Large Language Models are becoming highly efficient at de-anonymizing internet users. By analyzing linguistic patterns, these models can link pseudonymous accounts to real identities with 85% accuracy. This process does not rely on leaked databases or IP addresses. It focuses entirely on the unique way you construct sentences and use specific vocabulary across different platforms.

The era of hiding behind a screen name is effectively over because your writing style is a biometric marker. A model can scan millions of posts to find a match between an anonymous whistleblower and a public profile. This capability transforms stylometry from a niche forensic tool into a scalable method of mass surveillance.Time to rethink digital privacy when our own habits of expression become the very data points that betray us.

🧠 LLMs identify users by matching unique linguistic fingerprints.
⚡ The accuracy rate for identifying individuals across platforms is 85%.
🎓 Anonymity now requires actively masking your natural prose.
🔍 Automated deanonymization poses a direct threat to journalists and whistleblowers.

https://arstechnica.com/security/2026/03/llms-can-unmask-pseudonymous-users-at-scale-with-surprising-accuracy/
#Privacy #Cybersecurity #AI #DataProtection #security #cloud #infosec

Anthropic shared an experiment where sixteen Claude agents worked together to build a functional C compiler in Rust. This was not a single model following a script, but a team of autonomous agents using a shared repository to coordinate their work. Over two weeks and 2,000 sessions, they produced 100,000 lines of code capable of compiling a Linux 6.9 kernel.

The significance of this project is not the compiler itself, which still relies on external tools for certain tasks. The real lesson is in the infrastructure required to let AI work in parallel without human oversight. By using Docker containers and simple lock files, the agents managed their own merge conflicts and task distribution. This suggests that the future of development is less about writing individual lines of code and more about designing the environments where agent teams can operate.

🧠 Parallel agents completed a task that usually takes human teams months.
⚡ The $20,000 cost highlights the trade-off between compute and human labor.
🎓 Success depended on automated testing and feedback loops rather than direct instruction.
🔍 The resulting compiler successfully ran Doom and compiled complex open-source projects.

https://arstechnica.com/ai/2026/02/sixteen-claude-ai-agents-working-together-created-a-new-c-compiler/
#AI #Engineering #FutureOfWork #Anthropic #SoftwareDev #Dev #Software

Φ In the Phaedrus, Plato argued that the invention of writing would destroy our memory and replace true wisdom with a mere shadow of it. He believed that when we stop internalizing knowledge and start relying on external tools, we lose the ability to actually think. Thousands of years later, we are having the exact same conversation about Large Language Models and ChatGPT.

The danger of AI is not that it will become too smart, but that it will make us too lazy to be wise. True education is what Plato called a turning of the soul, a difficult process that requires active engagement. If you let a machine summarize the world for you, you are only holding onto dead speech. We must treat writing and thinking as a practice of the mind rather than a task to be automated.

🧠 Plato feared that external tools create the illusion of knowledge.
⚡ Large Language Models offer quick results while bypassing understanding.
🎓 Genuine insight comes from human dialectic and struggle.
🔍 We must focus on literacy that teaches how these algorithms function.

https://www.templeton.org/news/plato-warned-us-about-chatgpt-and-told-us-what-to-do-about-it
#ArtificialIntelligence #Philosophy #Learning #ChatGPT #Education #Teaching #AI #Technology

Regulators in Europe are finally calling out the mechanics of the digital slot machine. They are demanding that TikTok remove design features that keep users stuck in a loop. It is a significant move because it focuses on how a product is built rather than just what content it shows.

Most of us think we have poor willpower when we cannot stop scrolling. You are actually fighting a system built to keep you from leaving. The EU is forcing a change that might finally give your attention span a fighting chance.

🧠 Europe wants an end to the infinite scroll.
⚡ The mandate requires a redesign of the For You feed.
🎓 Policy is shifting toward protecting human cognitive limits.
🔍 ByteDance faces heavy fines if they ignore this order.

https://arstechnica.com/tech-policy/2026/02/eu-says-tiktok-needs-to-drop-addictive-design/
#DesignEthics #TikTok #SocialMedia #security #privacy #cloud #infosec #cybersecurity

Often, we entrust a single company with the keys to our entire lives. Conduent is one of those companies and they manage the vast digital paperwork for government agencies and state payrolls. This week they lost 12 billion records. It is the largest data failure in history, and we built a system with a single point of failure. Thieves now have the data to file for your tax refund or steal your benefits. The volume of this leak means your identity is likely available for sale. You must act to protect your money before a criminal uses your name.

🧠 12 billion records leaked from government service portals.
⚡ Social Security numbers and bank details are now public.
🎓 Criminals use this data to intercept unemployment checks.
🥶 Freeze your Credit NOW to stop identity thieves from opening new loans‼️

https://gizmodo.com/likely-the-largest-breach-in-u-s-history-what-you-need-to-know-about-the-conduent-fiasco-2000725930
#DataBreach #Privacy #CyberSecurity #security #privacy #cloud #infosec #cybersecurity

Google is fixing a big problem that most people don't realize we have. Our current encryption relies on math that quantum computers will eventually break, very easily. To prevent future hackers from reading today's data, Google is deploying post-quantum cryptography within HTTPS certificates.

This change uses the ML-KEM algorithm to protect connections. By changing how keys work, Google protects your private information before quantum computers can access it. Security is about winning the race against a threat that does not exist yet. 🔐

🧠 Google is using ML-KEM to protect HTTPS certificates from future quantum attacks.
⚡ Engineers use hybrid key exchanges to maintain compatibility with older systems.
🎓 Post-quantum cryptography protects data stolen today from being read in a decade.
🔍 The shift helps define new global standards for internet privacy.

https://arstechnica.com/security/2026/02/google-is-using-clever-math-to-quantum-proof-https-certificates/
#QuantumSecurity #Google #Encryption #security #privacy #cloud #infosec #cybersecurity

AI is not magic.

#AI #ArtificialIntelligence #Magic #RoseColoredGlasses #magicalthinking

#twitter #x #lastweektonight #johnoliver #freespeech #socialresponsibility

https://youtu.be/p7ZG_xWYLzI?si=y4r0sd3wEnzbYLd9

I spent the evening as a part of the Evanta Chicago CIO Inner Circle. As a Gartner company event, the room was filled with leaders navigating a tough 2026 budget cycle. We moved past the honeymoon phase of new tech to discuss how to turn high expectations into measurable financial results, all centered around AI, of course.

The consensus was clear. We cannot rely on static plans when the market moves this fast. Success depends on our ability to shift resources toward tools that actually address business bottlenecks.

🧠 AI adoption needs a single owner to bridge the gap between IT and the board.
⚡ Pivot your strategy the moment a project stops showing value.
🎓 Operational resilience requires a better way to calculate risk.
🔍 Focus on finding revenue gains instead of just saving time.

https://www.evanta.com/cio/chicago/chicago-cio-inner-circle-9277
#Chicago #CIO #Gartner #Leadership #CTO #CISO #AI #security #privacy #cloud #infosec #cybersecurity #Evanta #Gartner
@RHR_International@mastodon.social

Are you waiting for an AI miracle? Big projects usually fail because they're trying to change too much at once. My latest Forbes article shows you a better, faster way...

Focus on micro-transformations. You identify and fix a single small bottleneck and see results in weeks. This builds the trust you need to move toward larger goals later.

🧠 Pick one simple manual task.
⚡ Use one specific model or product for it.
💡 Leverage AI tools you already have in your stack, like Google.
🎓 Recognize your savings immediately.
🔍 Move to the next one.
🔍 Expand to a second team once the first succeeds.

A surgical approach builds the confidence your team needs. You stop waiting for a miracle and start seeing results right away.

https://www.forbes.com/councils/forbestechcouncil/2026/02/24/micro-transformations-a-practical-path-to-ai-roi-in-business/

#AI #artificialintelligence #leadership #CIO #CTO #CISO #Caio #innovation #business #security #privacy #cloud #infosec #cybersecurity

@RHR_International@mastodon.social
@forbes@flipboard.com
@Forbes@newsie.social
@forbestechcncl@bird.makeup

😳

Watching humanoid robots take the stage at the Lunar New Year celebration in China 🇨🇳 means a shift in our relationship with robots. You might think of robots as stiff tools confined to factories. Buckle up... These machines show a level of coordination that mirrors our own biology. 🤯 The speed of development is incredible. It took decades to teach a machine to walk. We are now seeing them dance and interact in front of millions of people. This transition means the gap between high-level computation and physical reality is disappearing.

🧠 Humanoid models now use neural networks for movement.
⚡ China plans to lead global production by 2026.
🎓 Performance art tests the limits of machine balance.
🔍 Industrial applications will follow these public displays.

https://www.reuters.com/business/media-telecom/chinas-humanoid-robots-ready-lunar-new-year-showtime-2026-02-16/
#FutureOfWork #Robotics #Engineering #Innovation #AI #security #privacy #cloud #infosec #cybersecurity

Have you ever asked AI one simple question…
after 3 seconds... you get impatient...

So you open another tab.
Ask another question to solve a different problem.
That takes 3 seconds... impatient...

Open another tab.... repeat 12 times...

By 5 p.m., you’ve got 27 AI answers, 14 browser tabs, and no idea what you were working on.

Welcome to my daily productivity strategy.

@boudah@hostux.social
“Palantir's software was "at the time the only tool existing on the market that met the needs of the DGSI to address national security concerns," it added. Already renewed twice, in 2019 and 2022, the contract has been extended this time "pending the deployment of a new sovereign tool," the DGSI said.”

💠 Holographic storage used to be the stuff of science fiction or expensive, failed experiments. We usually expect massive capacity at the cost of proprietary hardware that forces us to throw away everything we already built. HoloMem is taking a different path by making holographic tech work inside the tape libraries you already have in your data center. 😳 The move from magnetic tape to light-sensitive polymer ribbons is the core of this change. These cartridges look like standard LTO units but hold eleven times the data on a ribbon that is ten times shorter. Because the polymer is physically stable, you can turn off the air conditioning in your archive room and walk away for fifty years.

🧠 The system uses a five-dollar laser diode to write data as volumetric voxels.
⚡ One 200TB cartridge replaces more than ten LTO-10 tapes.
🎓 Software sees the HoloDrive as a standard LTO device for easy adoption.
🔍 Field trials in UK data centers show the tech works with existing robotic arms.

https://www.techtarget.com/searchstorage/news/366638887/HoloMem-demos-backward-compatible-holographic-storage
#Storage #DataCenter #TechTrends #security #privacy #cloud #infosec #cybersecurity #DataStorage #Archive #Backup #Tape #Storage #Data

🇫🇷 The French government is sending a message about digital independence by ditching Microsoft Teams and Zoom for its own platform, Visio. This move is not about a lack of features but a calculated push for digital sovereignty. By 2027, every government department in France will communicate via this sovereign stack, ensuring that sensitive data remains subject to European law. The shift addresses a growing concern that relying on foreign cloud giants creates a strategic vulnerability. France is opting for an open-source architecture built on Django and React, hosted on the domestic Outscale cloud. This infrastructure removes the risk of external service disruptions while fostering a local tech ecosystem.

🧠 The Visio platform integrates with Tchap, a secure messaging app based on the Matrix protocol.
⚡ French startup Pyannote provides the AI backbone for meeting transcripts and speaker identification.
🎓 The administration expects to save millions in annual licensing fees by using in-house software.
🔍 Data remains strictly within French jurisdiction to bypass foreign data access laws.

https://itsfoss.com/news/france-ditches-microsoft-teams-and-zoom/
#France #TechHistory #DigitalSovereignty #OpenSource #security #privacy #cloud #infosec #cybersecurity #FOSS

Prediction markets sell themselves as truth engines. Put money on outcomes, strip away noise, and let prices reveal reality. The Venezuela strike exposes the cost of that logic. When someone can profit from a covert military operation before the public knows it happened, markets stop forecasting the future and start rewarding proximity to power. This isn’t information discovery. It’s monetized access. The political problem is simple and corrosive: greed gets institutionalized. When insiders are encouraged to leak or act early because there’s money on the table, restraint becomes irrational. Journalism delays publication to protect lives. Markets pay you to move first and ask questions never. The ethical problem runs deeper. Prediction platforms don’t just reflect reality; they reshape incentives inside governments, corporations, and security institutions. They quietly ask every insider: do your job, or place your bet? And the security problem may be the most dangerous of all. Classified plans, military actions, and diplomatic moves become market signals. Prices move faster than accountability. Risk is no longer mitigated; it’s priced. Efficient markets are not moral systems. When everything becomes tradable, even secrecy turns into a commodity. And when insiders win, the public doesn’t gain clarity. It absorbs the risk. TL;DR 🧠 Prediction markets reward access to secrets ⚡ Greed displaces ethics and restraint 🎓 Security decisions become financial signals 🔍 Truth priced by markets is not the same as a public good https://www.theatlantic.com/technology/2026/01/venezuela-maduro-polymarket-prediction-markets/685526/ #MarketEthics #Politics #Crypto #Power #security #privacy #cloud #infosec #cybersecurity

This Gmail hack is unsettling not because it’s flashy, but because it’s bureaucratic. Attackers aren’t breaking encryption or outsmarting algorithms. They’re filling out forms. By changing an account’s age and abusing Google’s Family Link feature, they can quietly reclassify an adult user as a “child” and assume parental control. At that point, the rightful owner isn’t hacked so much as administratively erased.

The clever part is that everything happens inside legitimate features. Passwords are changed. Two-factor settings are altered. Recovery options are overwritten. And when the user tries to get back in, Google’s automated systems see a supervised child account and do exactly what they were designed to do: say no.

Google says it’s looking into the issue, which suggests this wasn’t how the system was supposed to work. But it’s a reminder of an old lesson. Security failures often happen when protective mechanisms are combined in ways no one quite imagined. The tools aren’t broken. The assumptions are.

There’s no dramatic fix here, only mildly annoying advice that suddenly feels urgent. Review recovery settings. Lock down account changes. Use passkeys. Because once an attacker controls the recovery layer, proving you’re you can become surprisingly difficult.

TL;DR
🧠 Family safety tools are being weaponized
⚡ Account recovery can be shut down entirely
🎓 Legitimate features enable the lockout
🔍 Prevention matters more than appeals

https://www.forbes.com/sites/daveywinder/2025/12/07/google-looking-into-gmail-hack-locking-users-out-with-no-recovery

#Cybersecurity #Gmail #IdentitySecurity #AccountRecovery #DigitalRisk #security #privacy #cloud #infosec