Is AI Dangerous? Ex-Employees Speak Out.
Hey there.
Windows sparked controversy with a new 'recall' feature that screenshots everything you do. OpenAI continues to attract negative attention, with employees signing a letter to warn about questionable practices. In personal development, we'll be learning the crucial skill of saying no. Let's dive in but first, a shoutout...
In Today’s Edition:
Window’s AI Feature Raises Security Concerns
100 Robots Serve Coffee at This Starbucks in South Korea
CustomGPT’s Are Now Available to Free Users
The Art of Saying No
Tools of the Week and more…
📢 Shoutout
TLDR covers the best tech, startup, and coding stories in a quick email that takes 5 minutes to read.
No politics, sports, or weather.
Subscribe for free now and you'll get our next newsletter tomorrow morning.
PS: Showcase your projects and ideas to over 2500+ readers every week, contact me directly on X/Twitter for more details.
🖥️ Tech Bytes
Window’s AI Feature Screenshots of Everything Raises Security Concerns
Microsoft plans to launch an AI-powered feature, Recall, for Copilot Plus PCs on June 18th. This feature provides users with quick access to their digital activity by taking screenshots of every user interaction on the PC and allowing for easy retrieval through a search interface. While Microsoft claims that the data will be securely encrypted on the device, cybersecurity expert Kevin Beaumont has raised concerns that the screenshots are stored in plain text format, making them vulnerable to exploitation by malicious actors.
He has also demonstrated automated methods to exploit this vulnerability, despite Microsoft's assurance that remote exfiltration of Recall activity is impossible. Additionally, Recall is set to be activated by default on new Copilot Plus PCs, with the opt-out option buried deep within the setup process.
Although Microsoft insists that Recall is an optional feature with strong privacy controls, Beaumont's findings point to potential security gaps, especially in relation to malware targeting the Recall database.
100 Robots Serve Coffee at This Starbucks in South Korea
A South Korean Starbucks has gained attention for its high-tech approach: 100 robots are now delivering coffee! While only two human baristas are in charge of making coffee, these robots handle the service, moving around the cafe to bring your order directly to you, no matter where you're sitting.
Developed by Naver Labs, a South Korean tech company, these robots showcase the increasing use of automation. Special software monitors them closely, ensuring they stay on track and deliver your coffee efficiently. Just like Amazon's large robot workforce in their warehouses, this cafe illustrates how robots are increasingly involved in everyday tasks.
This robotic Starbucks isn't the only example of automation becoming more prevalent. Companies such as Naver, Tesla, and Engineered Arts are leading the way in this technological shift, shaping the future of how we work and live.
OpenAI, DeepMind Accused of Downplaying AI Dangers
A group of current and former employees at leading AI companies OpenAI and Google DeepMind published a letter on Tuesday. They warn against the dangers of advanced AI, alleging that these companies prioritize financial gains and rush to develop increasingly complex systems while avoiding oversight and downplaying potential risks.
The letter, titled "A Right to Warn about Advanced Artificial Intelligence," is signed by thirteen current or former employees of OpenAI. Two signatories are from Google DeepMind, and six anonymous individuals also express concerns about potential backlash.
The researchers have made four demands of AI companies: stop using non-disclosure agreements to silence employees, establish a method for employees to report safety concerns anonymously, foster a culture of open criticism, and ensure no retaliation against whistleblowers.
Daniel Kokotajlo, a former employee at OpenAI, has published a thread on X detailing his concerns regarding the issue. Furthermore, Anthropic is also advocating against AI regulation.
CustomGPT's Are Now Available to Free Users
Free users of ChatGPT got a significant upgrade in May with the addition of features previously exclusive to paid plans. The advanced GPT-4o model powers these new features and allows users to explore custom-built versions of ChatGPT, analyze data, create charts, and even ask questions about pictures. However, there are limitations for free users. They cannot create their own custom GPTs and will be switched back to the older GPT-3.5 model after reaching a certain limit of messages with GPT-4o. Additionally, the controversial Sky voice option is no longer available on any tier of ChatGPT.
🌱 Personal Growth
The Art of Saying No
Feeling overwhelmed with requests and tasks? Saying no might seem counterintuitive, but it's the secret weapon you need to focus and achieve your goals. Knowing what to prioritize and having the confidence to say no is a vital skill. However, learning how to say "No" effectively is even more crucial. Find out how this skill can help you excel easily.
🎁 Tools of the Week
Otter.ai - A video assistant that records and transcribes meetings.
Letsrecast: AI tool to create engaging podcasts of your favorite articles.
Getreadytosend: Auto-generates replies for Gmail before you open the email.
These were my top insights from this week. Stay tuned for more content on tech, AI, no-code, and entrepreneurship. If you liked this edition, do let me know on X/Twitter. I also share video content on my Instagram page.