Edison Figueira @EdisonFigueira - Twitter Profile

14 days ago

I’ve been on Linux since SLS. Red Hat. RHCE. Then Debian and Ubuntu. I still kept a Mac. Not because I disliked Linux — because video froze, audio died, and vendors barely help. The Linux desktop also never quite had an identity. Not Mac. Not Windows. Just… Linux. Then I tried Omarchy. Different. Easy where it counts. The basics actually work. It has a face of its own. AI feels built in, not bolted on. First time in years a Linux desktop hasn’t felt like a compromise. Hats off to the project and to DHH. @dhh

0

2

0

31

EdisonFigueira retweeted

Nav Toor

@heynavtoor

6 months ago

🚨BREAKING: The man who won the Nobel Prize for inventing modern AI just said he's "more worried than ever." Geoffrey Hinton quit Google. Warned Congress. Told the world AI changes everything faster than anyone expects. But buried in his 40 years of research is a reasoning framework 99.9% of people have never seen. Here are 9 Claude prompts built on Hinton's neural architecture that turn Claude from a chatbot into a deep reasoning engine:

heynavtoor's tweet photo. 🚨BREAKING: The man who won the Nobel Prize for inventing modern AI just said he's "more worried than ever."

Geoffrey Hinton quit Google. Warned Congress. Told the world AI changes everything faster than anyone expects.

But buried in his 40 years of research is a reasoning framework 99.9% of people have never seen.

Here are 9 Claude prompts built on Hinton's neural architecture that turn Claude from a chatbot into a deep reasoning engine:

65

800

134

1K

118K

EdisonFigueira retweeted

Google AI Developers

@googleaidevs

6 months ago

Learn how to fine-tune FunctionGemma on TPUs in a @GoogleColab notebook. This guide uses Tunix, a lightweight JAX library, to streamline post-training your LLM🧵👇

5

317

55

181

21K

EdisonFigueira retweeted

Rohan Paul

@rohanpaul_ai

7 months ago

Yann LeCun (@ylecun ) explains why LLMs are so limited in terms of real-world intelligence. Says the biggest LLM is trained on about 30 trillion words, which is roughly 10 to the power 14 bytes of text. That sounds huge, but a 4 year old who has been awake about 16,000 hours has also taken in about 10 to the power 14 bytes through the eyes alone. So a small child has already seen as much raw data as the largest LLM has read. But the child’s data is visual, continuous, noisy, and tied to actions: gravity, objects falling, hands grabbing, people moving, cause and effect. From this, the child builds an internal “world model” and intuitive physics, and can learn new tasks like loading a dishwasher from a handful of demonstrations. LLMs only see disconnected text and are trained just to predict the next token. So they get very good at symbol patterns, exams, and code, but they lack grounded physical understanding, real common sense, and efficient learning from a few messy real-world experiences. --- From 'Pioneer Works' YT channel (link in comment)