🤝 Next-Gen Tools to Revolutionize Your Tech Experience
Dear curious minds,
Welcome to another exciting edition of my newsletter! This year's Consumer Electronics Show (CES) is a whirlwind of innovation and excitement, but there's one clear standout: AI has taken center stage, transforming the tech landscape and captivating our imaginations. I will not cover all the announcements from the CES, but one new gadget did make it in this issue. Besides that, there is again exiting news from OpenAI and another cool tool I discovered.
In this week's issue, I bring to you the following topics:
ChatGPT Team: A Privacy-friendly Subscription Plan
GPT Store: A First Look Beyond the Hype
Rabbit's r1 Aims to Make Smartphones Obsolete
Fixkey: Typing at the Speed of Thought on macOS
If nothing sparks your interest, feel free to move on, otherwise, let us dive in!
Thanks for reading! If you are not already subscribed, enter your mail address below to receive new issues and support my work.
🔒🤝 ChatGPT Team: A Privacy-friendly Subscription Plan
OpenAI has launched ChatGPT Team, a subscription plan tailored for small businesses and teams. It's designed to provide advanced AI capabilities in a secure, collaborative workspace.
Access to GPT-4 with 32k context window (to be honest, I do not know what ChatGPT Plus subscribers get access to right now, announced was already a context window of 128k token)
Higher message cap for GPT-4
No training on your data
Team workspace to easily share custom GPTs
ChatGPT Team costs $25/month per user (billed annually) or $30/month per user (billed monthly) and needs at least two team members. With that, it is an attractive alternative to ChatGPT Enterprise which costs $60/month and requires at least 250 users and needs to be paid upfront for a complete year, resulting in an investment of at least $180k.
My take: One of the most significant advantages of the Team plan is its commitment to privacy. In contrast to the Plus plan, where user data may be utilized for future AI training, the Team plan assures users that their interactions will not be used for such purposes. This feature is crucial for those prioritizing data privacy. You might argue that you can also turn the privacy-mode on in the settings as a ChatGPT Plus user, but you will lose your history and advanced features like custom GPTs.
🚀🔍 GPT Store: A First Look Beyond the Hype
There's been quite a buzz around OpenAI's GPT Store, and I already covered GPTs in general and the announcement of the upcoming store release in last week’s issue.
The launch article states that OpenAI highlights every week different GPTs and furthermore, the most used GPTs in various categories are shown.
Contrary to what some might have thought, the GPT Store does currently not feature user ratings or the option to buy GPTs. It's more about discovery and sharing than a traditional marketplace.
To use the GPT Store, you need to have a paid ChatGPT subscription (Plus, Team or Enterprise).
OpenAI announced to introduce a revenue-sharing model, targeted to start in Q1. However, it's important to note that this is, for now, limited to users in the United States.
My take: Thousands of users did dive into building GPTs after the launch in November with the goal to monetize them in the announced GPT store. Likely there will only be a few lucky persons besides big companies with proprietary data or APIs which will get large payments.
🐇🥕 Rabbit's r1 Aims to Make Smartphones Obsolete
A company called Rabbit launched the r1, a dedicated AI voice assistant device. Fully AI-powered with natural language interface instead of apps. The goal is to help people by completing tasks through voice commands instead of using countless apps.
The Rabbit r1 was launched at the CES (Consumer Electronics Show) and you can watch the announcement video here.
They state that the r1 runs a LAM (large action model) which in contrast to a LLM is built to take action. It can do things like play music, order food, book trips end-to-end after connecting accounts. Furthermore, it can learn to complete tasks by watching you do them once.
Comes with a camera, microphone, speakers for computer vision features.
Works together with a web portal to control privacy and connected accounts.
Costs only $199 with no subscription fees and is announced to already ship around Easter 2024, but the first batch is already sold-out.
My take: While the Rabbit r1's demo appears smooth and its price tag attractive, it's crucial to address the privacy concerns inherent in such devices. The hardware, not potent enough to run models locally, and the r1 relies on cloud services. This dependency raises questions about data privacy and security. Users are required to trust that their data is handled appropriately, a concern that isn't unique to Rabbit r1 but is prevalent across AI-powered devices relying on cloud-based operations.
⌨️⚡️ Fixkey: Typing at the Speed of Thought on macOS
Forget the Copilot key, which was announced by Microsoft and covered in my last issue. What you really need to check out is Fixkey — a game changer which allows you to type at lightning speed and get the text corrected by a locally running LLM.
A native macOS application which supports over 200 languages and more than 300 apps.
You can test it in a free trial and need to subscribe to continue usage for $50 yearly or $6.90 monthly.
My take: Sadly only for macOS, but the concept is so good that I am confident there will be alternatives for other operating systems soon. I already started to discuss a Linux realization with GPT-4. Imagine having a physical key (or a mapping) for it! The only problem I see arising is that I will never be able to type a correct sentence without it. But jokes aside, Grammarly will get serious problems!
Disclaimer: This newsletter is written with the aid of AI. I use AI as an assistant to generate and optimize the text. However, the amount of AI used varies depending on the topic and the content. I always curate and edit the text myself to ensure quality and accuracy. The opinions and views expressed in this newsletter are my own and do not necessarily reflect those of the sources or the AI models.