Good morning to all our bright-minded readers, and a warm welcome to the refreshed edition of Blacklynx Brief. We stopped counting our editions, but since you’re wondering, this is volume 13.

We stopped counting because quite possibly, with the insane speed of development, we might be moving to two editions per week.

In any case, the air of transformation is buzzing through our pages. You might notice a splash of change, a dash of the new—yes, it seems we're in those formative years where, like a curious teen, we're shaking things up, dabbling in a little bit of everything.

Fear not, for this isn't your usual adolescent rebellion. Every tweak and turn in our style has been carefully crafted, inspired by the most important critic of all—you.

So, grab your coffee, and let's dive into the vibrant world of AI, cybersecurity, crypto, and technology. Yes, I will talk about crypto in the future.

Let’s get started!

[Cyber] [AI]
Joe Biden’s Executive Order on AI

The big news this week in both the cybersecurity and AI worlds was that in the United States the political world got involved with AI safety.

In a groundbreaking move, President Biden has unveiled a sweeping executive order on AI, which is so detailed we had to use ChatGPT to interpret and understand it. This pioneering measure is set to bring order to an industry that has, until now, been navigating without much oversight.

The major upshot? AI innovators will now be required to demonstrate the safety of their advanced technologies to the federal government, affirming they pose no danger to the public.

The initiative calls for a collaborative approach across government entities to prevent the technology from causing disruption.

Data privacy is also being addressed in the order, but it’s more about giving advice than actual directives.

The main criticism of this legislation is that, as companies are racing towards achieving superintelligence, strict regulation would give companies outside of the US, particularly China, an advantage in getting there first.

AI development is reminiscent of the race to develop nuclear weapons during World War II.

Whoever gets there first will rule the world.

[AI]
Microsoft Unleashes Copilot

Your company may soon embrace Microsoft Copilot, which has been integrated into the enterprise Microsoft 365 subscription. 

Copilot is a lot like Clippy on steroids, designed to streamline your workflow. It promises to convert Word files into dynamic PowerPoint slides, craft emails, sum up the key points from your video meetings, and handle a variety of other tasks across the Microsoft 365 platform.

Companies interested in leveraging Copilot's capabilities need to acquire at least 300 licenses, priced at €30 per user, which will be an addition to the standard Microsoft 365 E3 or E5 subscription costs of €36 and €57 per user per month.

Microsoft's pioneering move to infuse ChatGPT into one of the most prevalent office software suites could be a watershed moment. Some speculate whether this integration could be Microsoft's equivalent to the launch of the iPhone. This collaboration with AI could potentially navigate us through our daily office tasks and could see Microsoft reaping as much as $10 billion in yearly revenue by 2026.

Select high-earning companies in the U.S. had the opportunity to trial Copilot earlier, and reports from one such company, as covered by the Wall Street Journal, indicate that employees saved between five and ten hours each month thanks to the tool.

However, one key insight from the past year's experience with ChatGPT is that while AI is a powerful tool, it is not infallible. Therefore, it's wise to review the AI's output for any irregularities before finalizing any critical communication.

[Cyber]
The week in cybersecurity

SolarWinds Scandal Explodes: The SEC has charged software giant SolarWinds and its security chief with fraud, revealing a shocking disregard for cybersecurity that left investors in the dark. Despite knowing their cybersecurity was lacking, the company falsely assured the public otherwise.

German Government Paralysis: A massive ransomware strike has thrown local government systems across Germany into chaos, targeting the IT infrastructure of the nation's heartland, North Rhine-Westphalia.

Russia's Payment System Under Fire: The NSPK, Russia's backbone for national payment processing (Russia’s SWIFT counterpart), has suffered a serious breach. We wouldn’t be surprised Ukraine is behind the attack.

The British Library Blackout: A mysterious cyber-attack has crippled the British Library's IT network, disrupting services for days.

Meta's Privacy Price Tag in the EU: In a groundbreaking move, Meta is rolling out a subscription model in Europe to sidestep GDPR ad targeting restrictions. Users must now weigh the cost of privacy against the perks of a free service, setting a precedent that could redefine the business of personal data in the EU.

Canada's Crackdown on WeChat and Kaspersky: Citing severe security risks, Canadian officials have barred government workers from using WeChat and Kaspersky apps on state devices. This bold move amplifies the global debate over the fine line between tech convenience and national security.

Facial Recognition Under Federal Scrutiny: In an era where privacy faces off against security, House Democrats push forward the Facial Recognition Act, aiming to rein in law enforcement's use of the controversial technology. As the bill demands warrants for facial ID, a national conversation on surveillance sparks anew.

[AI]
The week in AI

ChatGPT’s Leap: The big news this week in the AI world - beyond the executive order was that ChatGPT now added the ability to upload any document and “chat” with it. Want to perform data analysis - just throw in your Excel and start asking questions. On top of that you don’t need to switch between tools anymore. Just ask for a picture for example and you’ll get one.

This also kills off many startups at the same time - unfortunately. If you were developing a plugin for ChatGPT that interprets PDF files or interprets X-Rays - you’re basically toast.

Silicon Whispers of GPT-4-2: Github Copilot , which is a tool that helps coders to program quicker - is going to adopt OpenAIs model GPT4.2 - which was specifically designed for coders.

MetNet-3 Unleashed by Google: MetNet-3 is an advanced weather forecasting model by Google that outshines traditional physics-based models with precise 24-hour forecasts. It uniquely uses direct atmospheric observations, featuring a new densification technique for integrating sparse data sources like weather stations into detailed predictions with 2-minute and 1-4 km resolutions.

It predicts weather variables like precipitation and wind more accurately than existing models and produces probabilistic forecasts. Google has implemented MetNet-3 to deliver real-time, high-resolution weather information across its products, enhancing weather-related decision-making for users.

As I’m writing this - the building I’m sitting in is being battered by Storm Ciaran. Which seems a lot more severe than predicted.

Voice of Health: Canadian researchers have trained an AI model to diagnose type 2 diabetes by listening to a patient's voice for six to 10 seconds. This machine-learning AI identified 14 vocal differences between individuals with and without the condition, including subtle changes in pitch and intensity that are undetectable to the human ear. By also considering basic health data like age, sex, height, and weight, the researchers believe this method will significantly reduce diagnosis costs.

Google’s Billion-Dollar AI Play: Google, in a strategic power move, infuses billions into Anthropic, casting a resounding vote of confidence in the AI startup’s potential. With Claude, Anthropic’s flagship AI, standing as a formidable rival to OpenAI’s offerings, this investment marks a significant escalation in the high-stakes AI dominance race.

Jimmy Apples' Bold Claim: The enigmatic Jimmy Apples stirs the pot of AI speculation, asserting that a ‘median-human’ level AGI has been attained. With AI alignment seemingly in check, the revelation catapults the discourse on automation's impact on the job market into the spotlight, igniting debates on the future interplay between artificial intelligence and human employment.

[Closing Thoughts]
What We Learned

Well, it’s been another interesting one. As a Microsoft customer I’m pretty disappointed not being able to use Copilot for the time being. On the other hand , 30€ a month per user is on the steep side.

The ChatGPT update on the other hand is going to be a massive boost in productivity. It’s also going to kill a lot of startups :(

Until next week, keep exploring and stay curious!