The Blacklynx Brief - Vol.8 - A HUGE Week

Welcome to Volume 8 of the Blacklynx Brief

Is it just me, or since I started this newsletter, has the world of AI and cybersecurity decided to dance a little closer? I swear, every time I glance sideways, AI’s picked up speed like a caffeinated squirrel.

This week? It's a smorgasbord of significant happenings. And by the way, introducing a shiny new section this week: the "AI Tools & GPT Gimmicks" corner.

Here, I'll spotlight an AI tool or a cheeky ChatGPT hack that might just sprinkle a bit of magic on your mundane. Today's feature? An AI tool that crafts automated offers for small businesses.

Pour yourself a coffee, maybe two..

It’s a long one ..

Cybersecurity News

Signal Foundation Adds-Post Quantum Cryptography

You might have heard that quantum computing is a huge threat to cybersecurity in general because quantum computers are able to break encryption algorithms that would take literally lightyears for a “regular” computer to break. The foundation behind Signal - one of the most secure messaging apps in the business and used by let’s say a “diverse clientele”. PQC or post quantum cryptography is a new field in cryptography thats adds additional layers of security to protect against quantum computers breaking your encryption.
Today’s encryption protocols are doomed to die and Signal is one of the first companies that is tackling this issue.

Another set of Authors are suing OpenAI over Copyright Infringement

In a class action complaint against OpenAI, authors of a wide range of fictional works are requesting a permanent injunction to prevent alleged copyright infringement. Part of the lawsuit some heavy hitters including George RR Martin (who should be writing the next “Game of Thrones” novel instead of suing people) and John Grisham. Other famous authors : David Baldacci, Jonathan Franzen, Michael Connolly, Jodi Picault, ..

In addition to breaking copyright laws, the plaintiffs claim that training large language models (LLMs) on illicit copies of writers' works amounts to "systematic theft on a mass scale." The complaint claims that "Defendants would have no commercial product with which to damage—if not usurp—the market for these professional authors' works if Plaintiffs' copyrighted works were not available to 'train' their LLMs. This is unfair and perverse." As a result of the defendants' deliberate copying, the plaintiffs' works become tools of their own demise.

International Crime Court Suffers Cyberattack

The International Criminal Court (ICC) at The Hague recently faced a cyberattack and is now working with Dutch authorities to investigate the incident. Interestingly, this breach came around the same time the ICC announced it would be considering cases of significant cybercrimes. Experts highlight that while the ICC is moving to strengthen their cybersecurity and consider cloud services, the cloud alone isn't a full security solution. However, it can aid in implementing better security measures. For a deeper dive, check out reports from Dark Reading, Bleeping Computer, and others.

USBs : An Old Threat Resurfaces

At the mWise security conference, Mandiant reported a resurgence in the use of USB drives by hacking group UNC53, linked to China, to disseminate malware. Since January 2022, they've breached at least 29 organizations with the Sogu malware variant. While technology evolves, old tactics remain effective. Ensure your security systems, especially newer EDRs (Endpoint Detection & Response) , are set to optimal protection. Learn more at Wired.

Ransomware Attack Leads to Downfall of UK Transport firm

KNP Logistics Group, once a parent to the historic 158-year-old Knights of Old haulage firm, has entered administration following a substantial ransomware attack in June that disrupted pivotal systems and financial records. They confirmed that roughly 730 workers would face redundancy, attributing the firm's compromised financial state to the cyber onslaught. Formed in 2016 from various mergers, KNP's vast history traces back to Knights of Old's beginnings with a single horse and cart in 1865. Most of the workforce will be let go, retaining only a few to aid in concluding its operations.

“All of Sony” Hit By Ransomware Attack ?

A recently-emerged dark web hacking group, Ransomed.vc, claims to have successfully compromised all of Sony's systems, according to a report by Cyber Security Connect on September 25. This new hacking outfit, believed to have ties with older dark web groups, allegedly uncovered 6,000 files from Sony, showcasing internal login pages, presentations, Java files, and much more. Interestingly, rather than ransoming the data, the group declared its intention to sell it. Ransomed.vc, purportedly operating from Russia and Ukraine, not only conducts ransomware attacks but also reports system vulnerabilities under the EU's data protection laws. Responding to the situation, Sony stated that they are currently investigating the matter.

Cisco to Acquire Splunk

Cisco has announced plans to acquire data analysis software giant, Splunk, in its most significant purchase to date. Both firms have greenlighted the acquisition, but it awaits regulatory approval. For a detailed read, explore insights from Cisco's official statement and coverage by major publications like TechCrunch and NY Times.

On to the AI news .. still with us ?

AI News

Microsoft Announces the Release of Copilot

Microsoft unveils its AI initiative, Microsoft Copilot, an effort to merge chat interfaces with large language models for a more interactive tech experience.

Copilot, designed to be a daily AI companion, is integrated into core products such as Windows 11, Microsoft 365, Edge, and Bing. It offers contextual assistance, emphasizing user privacy.

Starting September 26, it will be available in Windows 11 through a free update and will soon extend to other platforms. Alongside, Windows 11 will receive over 150 new AI-centric features. Additionally, Bing is set to offer an enhanced experience using OpenAI's DALL·E 3 model.

OpenAI : ChatGPT Voice + Browser Support

Did i mention that things are heating up? The vision of OpenAi is to evolve towards an everything to everything model. Where you can throw in an image and receive back audio. After the announcement DALL-E 3 would be included into the ChatGPT experience for Plus and Enterprise users, this week OpenAi announced it will allow its users to have voice conversations with ChatGPT akin to Apple’s Siri.

But whereas Siri has only preprogrammed responses these responses can be anything. Subscribers will also soon be able to send images to the chatbot for analysis. You could request a recipe list based on the picture of the actual dish. Or you could take a picture of your math homework and ask ChatGPT to solve it.

OpenAI also announced that it’s rolling out web browsing for the Plus and Enterprise clients (they’re making it very difficult for us to say no to a Plus account).

GPT Plus accounts are becoming irresistible

So from now on - no more disclaimers that the information is not up to date. ChatGPT can now surf the web - which is going to fuel the fears of the people who believe this is going to lead to Matrix-style robot overlords.

I checked and it’s already deployed

Rumours are swelling : is AGI already here ?

This is not news but a rumour. There are some unconfirmed speculations about internal OpenAI models. According to an unconfirmed leak OpenAI is working on an LLM that is codenamed “Arrakis”. Which is to be honest a weird name - are they suggesting that earth is going to be turned into a desert planet with sandworms burrowing through the core, like its namesake in the Dune saga ?

This is one of the “any-to-any” models we previously mentioned

Enter the rabbithole here

If you’re in the mood for a deep dive on AGI and why it’s going to be achieved quicker than you think : here’s an EXCELLENT article 

Amazon is dropping 4 Billion Dollars into Anthropic AI. Intel Makes 6 Announcements

The AI fight amongst the big tech giant is heating up even more. Billions of dollars are being deployed to gain an edge in the race. And this is in the US alone, we’re not even talking about China.

CHINA!

Amazon is doubling down on Anthropic, the studio that is building the Claude 2 model and other AI solutions for Amazon.

At their third annual innovation event Intel came out with six different announcements for the rest of the year : from new hardware to plans to building an AI supercomputer for Stability AI (Stable Diffusion).

Is AI killing the Internet ?

Something to consider. Another rabbit hole of which I will just show you the entrance 😀

Click here.

AI Tools & GPT Gimmicks

You know, as an aside, a few of you have contacted me and asked me to highlight some cool AI uses cases or tools I have come across.

Here’s one from Belgian soil to start things off.

Check out this awesome looking tool from Belgian-based AI studio Blackbear AI labs :

What they have built is quite impressive (Video is in Dutch) : they have ChatGPT go through all your previous sales offers and then then when you have a new prospect they will build the sales offer for that prospect. Pretty cool to see in action.

That’s it from us !

If you have learned something new this week - please pay it forward by performing a simple share on LinkedIn. Or send it to your friends.

Once again I’m not asking for your soul - just a little share 🙂 Until next week !

Reply

or to participate.