The Sunday recap✨

Howdy, wizards.

⏪ This is your weekly recap email on Sundays. All the best links I’ve shared during the week. No fluff or unnecessary details included.

🤷🏻‍♂️ Don’t need the Sunday recap? You can update your subscriber preferences at the end of this email to pick what emails you’d like to receive.

Let’s recap!

THE SUNDAY RECAP

THE MOST IMPORTANT DEVELOPMENTS IN AI THIS WEEK

1. Google’s AI-powered Chess is really cool

Google released GenChess, an experimental chess game that lets you customize your chess pieces using GenAI. You type in a description and it generates pieces using the Imagen 3 model. It also makes suggestions for a suitable opponent.

Google also has four other games powered by AI that you can try out right now:

• Food Mood: A recipe generator to inspire your next meal.

• Learn About: AI that helps you learn any topic at your own pace.

• Say What You See: A game that helps you get better at image prompting.

• Musical Canvas: Make a drawing and generate a fitting soundtrack for it.

My bet is you’ll soon be able to generate your own character in nearly any game. It’s a small piece of the gaming experience, but AI makes a valuable contribution here.

‎ ‎→ ‎ Read the full newsletter here

TOGETHER WITH HUBSPOT

Start using the CRM that’s 100% free — with something for everyone.

HubSpot offers an intuitive, top-rated custom relationship management platform designed specifically for small businesses. With its user-friendly interface and robust features, you can effortlessly manage leads, track sales performance, and gain a deeper understanding of your customers. Plus, it allows you to store and manage up to 1,000,000 contacts with no limits on users or customer data, making it a perfect fit for businesses at any stage of growth.

Whether you're just starting out or scaling up, HubSpot's comprehensive set of tools will help you stay on top of customer conversations, close deals faster, and grow your business. Best of all, HubSpot’s CRM is completely free to use. It’s time to focus on what matters most: your customers. Let HubSpot handle the rest.

Try HubSpot

2. Claude launches custom styles, making it easy to generate text in your own tone of voice

Claude now has the option to select writing styles, right inside the Claude interface. You have pre-defined styles including normal, concise, explanatory and formal — but the best thing here is that you can create your own custom styles very easily. You just need to upload some samples of your writing (text files, PDFs, etc) and it will create a style around it.

Teaching chatbots to write like you is nothing new, but the ease with which you can create, access and manage these styles right inside Claude is super convenient. Definitely going to be using this one.

‎ ‎→ ‎ Read the full newsletter here

3. Eleven Labs launch their own “AI overviews”

Eleven Labs has finally taken notes from what Google is doing with “AI overviews” inside NotebookLM. The new feature, called GenFM, lets you turn any PDF, article, e-book, research paper, web link, etc. into a personal podcast with AI co-hosts – and it can do this in 32 languages. You also get a web link for these so you can share/listen to your podcasts on the web. Currently only available for iOS, but Android is coming soon.

NotebookLM just got a competitor. I did a short test and the voices sound very life-like, however the natural feel of the narration—and dynamics between the hosts—is nowhere near as engaging as NotebookLM is. The biggest advantage here is that ElevenLabs supports far more than languages than Google currently does.

‎ ‎→ ‎ Read the full newsletter here

4. Anthropic’s standard for connecting AI to different data sources

Anthropic launched the Model Context Protocol (MCP) – an open-source standard for connecting AI assistants to data sources, including content repositories, business tools, and development environments.

Even the best of models are constrained by their isolation from you or your company’s data. If you want it to access a data source directly it requires some kind of custom implementation, which makes it hard to build fully connected systems.

The new standard gives an easy way for developers to allow AIs to connect to their application’s data through MCP servers or build AI that connect to these applications. It’s also built with security in mind and no need to share API keys with LLM providers. Anthropic has already shared pre-built servers for systems like Google Drive, Slack and GitHub.

A challenge making many AI use cases cumbersome to implement is that they require custom integration with several data sources. Anthropic’s cleaner solution of replacing fragmented implementations of data sources with a single protocol/ecosystem is very alluring and might be a step towards truly context-aware AI.

‎ ‎→ ‎ Read the full newsletter here

5. Zoom’s rebrand as an AI-first company

Zoom had an insane growth trajectory during the pandemic, but has since seen strong decrease in growth driven by people returning to the office (lessening the needs for the product) as well as intense competition from collaboration platforms that companies already pay for offering similar features including Google Meet, Microsoft Teams and Slack.

Zoom’s new strategy is, in a nutshell, rebranding themselves into an “AI-first communications platform”, dropping the word video from their name and focusing on their Zoom AI companion.

A lot of the comments on this rebrand go along the lines of “Zoom is simply slapping AI on everything”. For example, the big focus project is the upcoming AI avatars feature that can record and send messages to your team, with the long-term vision of having AI clones of yourself attend meetings. There might be a version of the future where this becomes super popular, but also it kind of sounds like building something that “could’ve been an email”.

‎ ‎→ ‎ Read the full newsletter here

FROM OUR PARTNERS

Love Hacker News but don’t have the time to read it every day? Try TLDR’s free daily newsletter.

TLDR covers the best tech, startup, and coding stories in a quick email that takes 5 minutes to read. And it's read by over 1.2 million people!

Try TLDR

6. Anthropic gets $4 billion from Amazon, which takes aim at Nvidia

Amazon and Anthropic are deepening their partnership. Here’s the details:

• Amazon is pumping another $4 billion into Anthropic, bringing its total funding amount to $8 billion.

• Amazon Web Services (AWS) becoming Anthropic’s “primary training partner”. They’ll use Trainium and Inferentia chips for future models; this ties directly into Amazon’s plan to rival Nvidia with its own AI chips.

• Anthropic’s AI, Claude, will power the new “intelligent” Alexa voice assistant (coming in 2025). Claude did a better job when tested vs Amazon in-house model, which they initially planned to use.

The partnership has strong synergies:

• Anthropic is getting access to the financial resources and compute it needs to compete with OpenAI, Google and Meta in advancing state-of-the-art AI models.

• Amazon, on the the other hand, is at a make-or-break moment when it comes to loosening Nvidia’s complete grip of the $100-billion-plus market for AI chips. Anthropic will be customers, as well as actively developing, hardware and software for frontier model training.

‎ ‎→ ‎ Read the full newsletter here

7. Linkup is building a legal bridge between LLMs and content publishers

French startup Linkup is creating a solution to how publishers can monetize their content when readers get information directly from chatbots. Tools like ChatGPT (with web search), Perplexity, even Google itself with AI overviews, demonstrate just how quick and convenient AI is becoming at giving us up-to-date answers from around the web. In some instances they have a licensing deal in place, but often it’s scraped information that the publisher simply isn’t compensated for.

Linkup's API connects developers of AI tools (or rather, any tools that use LLMs) with premium content sources through direct licensing deals and CMS integration so they can fetch content without scraping. Publishers earn revenue based on how much their content is accessed by the applications.

Some say that the health of the web, in this age of AI companies scraping it all up without paying for it, can only be saved by new business models – this is potentially one such approach. It’s encouraging to see innovation like this that can potentially ensure the longevity of quality content on the web.

‎ ‎→ ‎ Read the full newsletter here

THAT’S ALL FOLKS!

Affiliate disclosure: To cover the cost of my email software and the time I spend writing this newsletter, I sometimes link to products and other newsletters. Please assume these are affiliate links. If you choose to subscribe to a newsletter or buy a product through any of my links then THANK YOU – it will make it possible for me to continue to do this.