Really cool features of the new Mac Studios by Apple: with just 20,000$ of hardware, you can run the FULL DeepSeek R1 model!
https://x.com/alexocheema/status/1899735281781411907?s=42
There’s no other company in the world that allows you to run R1 so cheaply!
With all this LLM news going around, we forgot the most exciting thing Apple 🍎 did this year.
They renewed Ted Lasso for Season 4 ⚽️ 🥅
https://www.hollywoodreporter.com/tv/tv-news/ted-lasso-renewed-season-4-apple-tv-1236163347/
Called it
“Perhaps surprisingly, European banks have emerged as a hotbed of experimentation. Strict confidentiality rules limit how much the financial sector can rely on cloud-based ai services. That makes open-source models hosted internally an attractive alternative. Natwest and HSBC, two British lenders, are both experimenting with building their services on top of R1, as is Spain’s BBVA, according to the Information, a news site.”
https://www.economist.com/business/2025/03/13/western-companies-are-experimenting-with-deepseek
Big news in LLM land landed yesterday!
Allenai.org released a fully open-source LLM (data, code, weights, and details) model, OLMo 2 32B.
They claim it outperforms GPT 3.5 and GPT 4o mini while being relatively small. If you care about open models (why should you care? I’ve written an article about that https://articles.xebia.com/deepseeks-v3-r1-ai-smarter-cheaper-secure), you ought to try it out!
Blog post with the announcement: https://allenai.org/blog/olmo2-32B
Analytics at the speed of your thoughts!
MotherDuck and DuckDB just released a local UI for duckdb.
Work with notebooks, databases, and visualize the data all within the same interface.
I love it already—try it out with duckdb -ui
More info in the comments!
Small language models can be a game changer, but they have often to be trained to a very specific task!
That’s why we’re thrilled to host the How to train SLMs workshop at our Xebia office in Amsterdam!
We’ll have a hands-on training complete with a deployment walk-through!
https://lu.ma/5qrzjz3j?locale=en-GB
Anand Sahay Mayank Verma Preetpal Singh Adam Kawa Klaudia Wachnio (Zdunczyk)
What’s the impact of DeepSeek AI V3 and R1 on the hashtag#GenAI market and on coding in particular?
Yesterday, our CEO Anand Sahay asked for my perspective, so I set aside two hours in my agenda to find out. Little did I know back then that writing this post would have taken longer than coding a whole web application leveraging DeepSeek.
To form a perspective, I didn’t want to read what everyone was writing, but I set out to replicate with DeepSeek the application that I coded in ChatGPT in the summer of 2023 to send a reminder to our customers by SMS (see my blog post https://xebia.com/blog/how-i-learned-to-stop-worrying-and-love-llms/)
One of the application’s goals back then was to be portable so that every colleague, even salespeople, could use it.
However, I didn’t want Deepseek to rewrite it in Go: I wanted to learn something new, so I opted for an even easier app: a web page, all HTML and Javascript!
Lo and behold, after 15 minutes, I had a working web app (the only thing I had to update was the “Sender” field, as I didn’t specify it in the prompt!), as you can see in the screenshot.
It works with your MessageBird API key, it asks who sends the message, and what text you want to send.
Then, you need to upload a CVS file with names and phone numbers and it will send custom messages to your customers (in this case, the message was “Hi Giovanni, I’m testing deepseek-v3”).
And the most amazing thing: thanks to DeepSeek’s breakthrough technology, it was cheap, and you can run it on-premise or in the cloud (I opted for Fireworks.ai so that I could compare it with other models such as Mistral, Llama, and more).:
And, cherry on top, I could alternate using DeepSeek V3 and DeepSeek R1 (their reasoning model, which performs on par with the latest and greatest OpenAI model) to understand the reasons for certain choices so that I could correct it when necessary! Transparency is a big plus of DeepSeek when compared with the competition!
For the curious, here’s the prompt I used for the initial coding exercise: “Using the messagebird API, create an HTML page that allows me to input my messagebird API key, a CSV with names and phone numbers, and a text box in which I can enter the body of the SMS message that has to be sent via message bird to the phone numbers contained in the CSV file”
If you’re not leveraging hashtag#AI right now, you’re already behind. But here’s the twist—embracing AI without a clear strategy is just as dangerous. Missteps in AI integration lead to wasted resources, fractured teams, and even reputational damage. The stakes couldn’t be higher.
That’s why it’s my pleasure to join Claire Grosjean from Technology Credit Union (Tech CU) in an upcoming webinar to discuss
You will find the link in the comment to join us on March 4th at 1 PM ET.
Abhishek Bhuwania Hemant Ramnani Matthew S. Gosselin Kimberly Martin Klaudia Wachnio (Zdunczyk) Rob Dielemans Anand Sahay
I was not expecting X to grow their EBITDA from 13% to almost 50% under Elon, but here we are
Research shows that while 90% of executives view data as an opportunity 📈, 70% report minimal or no gains 📉.
The reason is that they miss a data strategy, meaning they basically have no clue where they’re heading. 🔙
So, where should they start?
My colleague Steven Nooijen shares 7 POWERFUL 🤜🏼🤛🏼 questions to get you started so you can play the game 🎮 WITHOUT losing money! 💰💰💰
Read the post on our blog!
