Alexis Rondeau

Bookmarks

5632 bookmarks · page 87 of 226

Bookmarked from Prajwal Tomar @PrajwalTomar_

Lovable is the only reason I’ve considered shifting from Next.js + Supabase to React + Supabase.

The speed is unreal. I can get 70-80% of the MVP up and running, deploy, gather client feedback, and then move to Cursor for enhancements and refinements.

This has completely transformed the way I build MVPs.

Bookmarked from ₕₐₘₚₜₒₙ @hamptonism

Every dark thing one falls into can be called an initiation. To be initiated into a thing means to go into it. The first step is generally falling into the dark place and usually appears in a dubious or negative form - falling into something, or being possessed by something... Even the worst things you fall into are an effort at initiation, for you are in something which belongs to you, and now you must get out of it.

Bookmarked from BOSS @thebeautyofsaas

Basic guide on picking career 2025 - going forward

Tier 1
Profit roles - Sales(ENT+ real high ticket), PE, M&A, HF, media buyers

Tier 2
Close to profit roles - Sales engineers, Business Analysts, CS roles specific roles in marketing

Tier 3
Rest

Don't sleep on this is crucial

Bookmarked from Euromaidan Press @EuromaidanPress

"Even the darkest night ends with dawn.”

This is how Ukrainian war veteran Yehor Babenko reflected on his year of recovery that tested his limits.

In March 2022, a border officer, who went to protect his country, suffered severe burns to his head, face, and hands during a Russian shelling in Mykolaiv Oblast.

He lost all ten fingers and could breathe only through a tracheostomy tube. Now, he awaits prosthetic fingers, an uncommon case due to the complete loss of all ten.

“The hardest thing is not to stop, even when it feels like the world around you has come to a halt,” Yehor wrote on Instagram.

He emphasized the importance of having support from family, friends, volunteers, and strangers. “Their faith in me helped me believe in myself,” he said.

“Thank you for having the courage to share your story,” wrote Ukrainian photographer Marta Syrko who took these photos.

📷martasyrko, egna77 Instagram

Bookmarked from Andrej Karpathy @karpathy

We have to take the LLMs to school.

When you open any textbook, you'll see three major types of information:

1. Background information / exposition. The meat of the textbook that explains concepts. As you attend over it, your brain is training on that data. This is equivalent to pretraining, where the model is reading the internet and accumulating background knowledge.

2. Worked problems with solutions. These are concrete examples of how an expert solves problems. They are demonstrations to be imitated. This is equivalent to supervised finetuning, where the model is finetuning on "ideal responses" for an Assistant, written by humans.

3. Practice problems. These are prompts to the student, usually without the solution, but always with the final answer. There are usually many, many of these at the end of each chapter. They are prompting the student to learn by trial & error - they have to try a bunch of stuff to get to the right answer. This is equivalent to reinforcement learning.

We've subjected LLMs to a ton of 1 and 2, but 3 is a nascent, emerging frontier. When we're creating datasets for LLMs, it's no different from writing textbooks for them, with these 3 types of data. They have to read, and they have to practice.