- The Blacklynx Brief
- Posts
- It's-a-me ! ChatGPT !
It's-a-me ! ChatGPT !
Banned in Italy!
WELCOME
Sometimes really significant events are being pushed out of the limelight by events that seem more spectacular and newsworthy.
Such a thing happened this week where the news was dominated with the farmer’s protests in the EU, and, oh by the way - Elon Musk’s Neuralink has implanted a chip into the first human brain.
It’s not only a noteworthy feat; it might be significant for humanity in general. It’s also something that can easily lead us to a horrific dystopian nightmare, but let’s worry about that when it comes closer and pretend the proverbial glass is half full.
I really enjoy writing this newsletter. The AI developments come hard and fast, but at the same time lawmakers are catching up with AI and for example, this week Italy slapped a ban on ChatGPT.
What a time to be alive !
AI NEWS
First Neuralink Chip Implanted in Human Brain
Microsoft's 'Future of Work' Report dives into how AI is changing our jobs, showing that AI tools like Copilot can make work better for 68% of people who use them. It points out that people in simpler jobs get a lot of help from AI, but there's a chance it could make the gap between workers bigger.
One big problem mentioned is relying too much on AI, which can lead to accepting wrong answers from it. Microsoft believes that making AI fit well into our daily jobs is just as important as the technology itself. This summary highlights why it's crucial to understand how AI is changing work now, to get ready for its bigger impact in the future.
OpenAI's CEO, Sam Altman, recently explored potential collaborations with Samsung and SK Group in South Korea to advance AI chip manufacturing, signaling OpenAI's interest in chip production. This strategic move, following discussions with Middle Eastern investors and Taiwanese manufacturers, aims to build a global AI chip production network. Such partnerships could safeguard OpenAI's computing needs, addressing Altman's concerns about future chip shortages and supporting the company's ambitious AI developments.
Researchers at the University of Cambridge have created a robotic fingertip that reads Braille at 315 words per minute with 87% accuracy, using AI to double the speed and accuracy of the best human readers. This system uses a camera to capture images and machine learning to sharpen and interpret the Braille letters quickly. This innovation is significant not just for its reading capabilities but for advancing touch-sensitive technology in robots, potentially enhancing their ability to perform complex tasks.
Elon Musk announced Neuralink's first successful human brain chip implant, with the patient showing positive signs of recovery. The chip, named 'Telepathy', aims to allow paralyzed individuals to control devices through thought, initially focusing on patients with spinal injuries or ALS. This milestone could revolutionize assistance for paralyzed patients and advance the technology towards broader use, marking a significant leap in brain-computer interface development.
This week, shares of AI giants like Microsoft, Google, AMD, and Nvidia dropped after their earnings reports didn't meet the high expectations set for the AI boom. Despite their efforts to lead in AI—like Microsoft's cloud and AI-driven Windows growth, and Nvidia's surging AI chip demand—increased spending to stay competitive has led to short-term financial pressure. This situation highlights the challenge of balancing the heavy costs of AI innovation with investor expectations for immediate returns.
AIWaves has launched Weaver, a new family of Large Language Models (LLMs) tailored for creative writing tasks such as fiction, blogging, and marketing. Unlike general LLMs, Weaver is trained on high-quality literary sources to master a wide range of creative styles. Its top model, Weaver Ultra, with 34 billion parameters, outperforms GPT-4 in creative writing benchmarks, highlighting its superior style, relevance, creativity, and fluency. This development addresses the need for specialized LLMs in creative fields, marking a significant step forward in AI-driven content creation.
CYBERSECURITY NEWS
Italy bans ChatGPT
Schneider Electric Ransomware Incident: Schneider Electric was hit by a ransomware attack on January 17, compromising its Resource Advisor cloud platform. The Cactus ransomware group, believed to have stolen terabytes of data, orchestrated the attack. This incident underscores the vulnerability of critical infrastructure to cyber threats.
Italy Finds ChatGPT Violates GDPR: Italy's data protection agency has determined that OpenAI's ChatGPT violates GDPR, imposing a temporary ban and demanding compliance measures within 30 days. This highlights the tension between AI development and privacy regulations in the EU. OpenAI faces potential fines if it fails to address these concerns.
Major IT Outage in Russia: A significant IT outage affected the Russian internet due to a DNSSEC error at the national domain registrar, likely stemming from attempts to establish a national DNS system. This incident reveals the complexities and risks of separating from the global internet infrastructure. It also highlights the potential for widespread disruption in a country's digital space.
Major BTC Seizure: German authorities seized over €2 billion in Bitcoin from a programmer associated with the defunct movie piracy portal movie2k.to. The voluntary turnover of funds and the domain control underscores the legal and financial consequences of early cryptocurrency adoption in illegal activities. This case also signals the growing effectiveness of law enforcement in tracing and securing digital assets.
23andMe Breach: Intruders accessed 23andMe customer accounts for five months, exploiting weak user credentials to steal sensitive genetic and health data. This breach highlights the critical importance of securing large databases of sensitive information against brute force attacks. It also raises questions about the responsibility of data holders to protect user information with robust security measures.
Ivanti Missed Patch Deadline: Ivanti acknowledged missing a deadline to release patches for vulnerabilities actively being exploited, citing the need for secure and high-quality fixes. This admission points to the challenges of maintaining security in legacy systems, especially following corporate acquisitions. It underscores the importance of timely vulnerability management in protecting against cyber threats.
PODCAST TIP OF THE WEEK
Darknet Diaries
The Darknet Diaries is a podcast about hackers, cybercriminals, fraud, and everything shady that is going on in the world.
Every week, it’s another horror story about exactly those things we, as information security professionals, are trying to protect companies and individuals from.
If you ever wonder what is going on on the dark web, this podcast will give you an idea.
Let me give you my top 5 favorite episodes to give you an idea of what this fantastic podcast is about:
1 : Episodes 45&46 : the Xbox Underground
This is a great one to get started; it’s a 2-parter about the Xbox hacking scene and how a group of guys pushed their luck a little too far. Extremely fascinating2 : Episode 36 : Jeremy from Marketing
A company hires a penetration tester to pose as a new hire, Jeremy from Marketing, to see how much he can hack in his first week on the job. It doesn’t go as planned.3: Episode 21: Black Duck Eggs
Ira Winkler’s specialty is assembling elite teams of special forces and intelligence officers to go after companies. Ira shares a story about a time he and his team broke into a global 5 company. A company so large that theft of intellectual property could result in billions of dollars of damage.4: Episode 72 : Bangladesh Bank Heist
This one is a bit special to me because I’m an extra in the story. I was working for one of the teams that was involved with the investigation at the time. Good times..
5: Episode 71 : Information Monopoly
This one is all about North Korea and how they have setup a formidable hacking force.
BOOK TIP OF THE WEEK
Peter Attia - Outlive
Peter Attia is a medical professional who specializes in "longevity.”. In his practice, patients come to him not because they’re sick but because they want to live as long as possible.
Everything he does centers around that goal. Last year, he wrote about his findings in this spectacular book.
The book is life-changing, as it is truly a wake-up call. It got me moving again.
What I especially like about the book is that it’s not written in a black-and-white style, like some health or diet books out there. You know the style, where diet is almost something religious. Attia takes a gentle and cautious approach.
This is the sign of the true master because he knows he doesn’t know a lot.
You have to become plant-based or you’ll die an early death, or keto is the only way, or paleo, or whatever. Lately, going meat-only is the new craze. None of this is in Attia’s book. He recognizes that everyone is different and that there is no one path to health.
I also like that he’s divulging a lot about his personal life and struggles in the book. Spoiler alert: It’s not pretty.
The book is centered around what Attia calls “The Four Horsemen of Death”. The four diseases that kill off most people : heart disease, cancer, neurogenerative disease (e.g Alzheimer's), diabetes type 2, and related metabolic syndrome.
He looks at what people who live to 100 have in common and how much of that is genetic. He devotes long chapters to the four horsemen and how they exactly kill us.
Then he looks at solutions on how to avoid getting killed. Of course, you can get unlucky and get a bad roll of the dice.
The gist of it is that exercise is the single most important factor that will allow you to live longer.
I’m doing a terrible disservice to the book now by trying to condense it in one sentence, but if you ever wanted an elaborate reason to get moving, this book is it.
It’s a fantastic read and highly recommended!
AI PRODUCTIVITY TIP OF THE WEEK
Microsoft Copilot
Two weeks ago Microsoft made their Copilot available for all organisations as they removed the 300 license minimum requirement.
It’s price point may seem rather steep initially , 30€ per user per month. I have not pulled the trigger myself yet but I hear that when your organisation has a lot of data dwelling on Sharepoint it becomes truly valuable as it trains itself on your data.
This data never leaves your company - although Microsoft is allegedly sending back “metadata”. Microsoft is notoriously beaconing towards their HQ all the time - so that’s not surprising.
Check out the video below on what you can do with Copilot.
Closing Thoughts
That’s another one in the books.
You might be getting some “ChatGPT-fatigue” - I know I have that at the moment, and it seems that AI development has hit a plateau, you cannot help but feel the cat is out of the bag and someone lost the bag.
Now .. let’s hope we can get the cat under control.
Thanks for being here and as always
please share this newsletter to a friend or colleague. Let’s keep this train moving forward !
Reply