Nothing scares a programmer more than this message 💀

My bets at the beginning of this year were these: ✅ Gemini will be the dominant model ✅ GitHub Copilot will be the default development tool And that the rest would gradually disappear or lose traction. Simply because of who has the data, the expertise, and the control within

Teslas have the lowest maintenance & repair costs of any brand

I couldn't be prouder of my new feature for creating images with my face 😄

How does an AI "read"? It doesn't see words, it sees "tokens". 🧩 Tokenization is the process of breaking text down into the smallest pieces a model can process. "fantastic" could be: - `fan` + `tas` + `tic` - `fantastic` Each model has its own tokenizer. That's why the same

I can't stop emphasizing how much you can learn by deploying your own services. A VPS on Hetzner, a Coolify installation, and off you go! 🚀

FUTURES TRADERS: Get 40% off all evals, no activation fees, end-of-day drawdown in our live-market PRO+ accounts…and still daily PRO payouts!

Could it be that I added the feature to create images with my face to the App to post on social media? Either that, or I went to the beach in the middle of October. One of the two.

Doesn't it feel like with these new plugins/extensions and skills in Claude and Gemini , we're addressing the same need but in different ways? 🤔 Wouldn't it have been better to just improve the MCP protocol? 🤷‍♂️ Am I missing something?

A developer can't just know how to code anymore. It wasn't enough before either, but you could hide it well because AI wasn't generating the code for you. A developer is starting to take on the role of a team lead, where AI can be an army of devs at their service.

Sometimes it doesn't matter what code the AI generated. With good architecture and solid tests, we can treat components as black boxes that we only peek into if something goes wrong. You were already doing this all the time with libraries.

Founded by the man behind Kellogg’s cereals, the Kellogg Foundation now pours billions into activism—not nutrition.

Cursor 2.0 is here, and it brings its own LLM! 🚀 New interface focused on AI agents instead of files ⚡ Introducing "Composer", their own AI model for code, 4 times faster and optimized for complex tasks. 🤖 The new multi-agent interface allows running multiple AIs in

Knowing how to define a good test suite will be the superpower of the coming years.

The foundation of software development was never about writing code. Code is a necessary byproduct, but irrelevant if we can move to a new level of abstraction. Developing software is about thinking, understanding the product, making architectural, technological, and

Does anyone know a good MCP that can control the Android UI, do screenshots and/or check the logcat? A Playwright equivalent for Android.

This is more than just four quarters. It’s every tailgate, every chant, every moment. It’s fuel that goes beyond the field. This is CELSIUS! LIVE. FIT. GO.

Last Thursday, I attended the accessibility Think Tank organized by @sngular, accompanied by @jcfmunoz and many other incredible professionals. I also finally had the chance to meet @JonOrcera in person! We discussed the challenges of integrating accessibility into company

As much as we're all hyped up about AI, if a problem can be solved deterministically, it'll always be 1000 times better than using an LLM.

@lmstudio 3️⃣ EXTRA: LM Studio can act as a server, and connect your Apps to your downloaded models via the OpenAI API. Perfect for developing prototypes without spending a single penny. Did you know about this? Do you use it often? Tell me in the comments 👇👇

@lmstudio 2️⃣ After this, you need to download the LLM you want to interact with. There are literally thousands, and it largely depends on your machine's power. I recommend trying a small but powerful one like Gemma 3n.

Premium men’s jewelry that elevates every fit. Black Friday deals live now!

1️⃣ Download @lmstudio, available for Windows, Mac, and Windows. Install it, and you'll find a chat interface that lets you communicate with multiple LLMs.

🚀 Run LLMs locally in 5 minutes 🚀 Did you know you can have AI locally, 100% private and offline in just a few steps? Today I'll show you how 🧵👇 (1/n)

I don't think we're adequately preparing for the radical shift happening in our software development industry.