I’m thrilled to announce that Lightning ⚡️ is now live on @Coinbase! I could not be more proud of the team for shipping it so quickly.

Growing Bitcoin adoption increases economic freedom in the world. This is my personal life mission and the mission for all of us at Coinbase.

5/ Link to the free Zora mint to own a digital collectible of this 15 month long investigation.

zora.co/collect/base:0…

In a Galton board, beads bounce off pegs as they fall with an equal chance of going left or right.

Beads accumulate in the middle as there are more paths that lead here, similar to how many random events in nature tend to cluster around an average.

🔴 VIDU: nuevo competidor de SORA

La empresa china Shengshu-ai presenta su modelo de generación de vídeo capaz de generar clips de 16 segundos en 1080p, con resultados próximos a Sora.

En su web no encuentro más info así que quedamos a la expectativa 👇

Different levels of light pollution

PROOF that Meta AI is sending your IP/location to the LLM in real time!

Within seconds of connecting to a new country via VPN, Llama 3 is already providing a greeting in the local language 😳

Why is this necessary? What other user info might Meta be sending to the LLM to

OpenVoice: Instant voice cloning.

Runs local on any computer.

Quality is quite good.

I’ll have a how-to soon.

YOUR AI I build in my garage will speak to you in 1000 voice based on situation.

Code: github.com/myshell-ai/Ope…

The most insane part of the Stripe announcement isn’t just that you can pay in crypto….

It’s that the SELLER receives normal fiat. Aka, they don’t need to know or care that you paid in crypto.

Your crypto stays with Stripe, then they pay out USD to the seller.

When

Delighted to release ✨Llama-3-8B-Web✨, the most capable agent built for web navigation by following instructions and replying💬. It surpasses GPT-4V* by 18% on WebLINX, a benchmark for web navigation with dialogue.
Model: huggingface.co/McGill-NLP/Lla…
Code: github.com/McGill-NLP/web…

Superfast Research Assistant using Llama3 on Groq Inc and tavily

Build an Assistant that:
🔎 researches a complex topic
✍️ writes a report at 800 tokens/sec

Try it yourself: git.new/groq-researcher

Here is the list...

Of SOME examples.

What does Chat GPT have to say on the subject?

The KING of YouTube:

Mr. Beast

Over the last 12 months, he's gained over 50,000,000 subscribers.

Just using one single trick: DUBBING

Here’s how you can do it too, in just 30 seconds:

AI Will Totally Transform Dating

Imagine an AI Agent that texts your Tinder matches on your behalf and then alerts you when the other party agrees to a date.

This could be a dream come true for most men, who have the right swipe pretty much everyone and then spend hours

The Real LLM SuperPower

LLMs have been excellent chatbots and coding assistants, and they have automated several tasks.

However, the real LLM superpowers are data cleaning and synthetic data generation!

These early LLMs help us create excellent, high-quality data, enabling the

If you are getting LLM fatigue, keep track of the following.

GPT-4 - best model for reasoning and code

Llama-3-70b - best open source alternative, 10x cheaper and faster

Llama-3-8b -a best small model for extremely fast inference and is good for custom fine-tunes

The Private LLM guys made Llama-3-Smaug from Abacus available on their iPhone app.

It's a very good small model... and works pretty well on a mobile device.

Who would have thunk it? - LLMs running on your phone 🤯🤯

Bindu Reddy A lot of people have underestimated Zuck and his ability to execute at an extremely high level at their own peril!

Apparently, SenseTime for China launched a model that beats GPT-4 on several metrics.

It's not open-source, so it doesn't matter all that much

That being said, I am hoping the fearmongers will shut up about China somehow not being able to access this tech 🙄🙄

Epic sat sold for 33 BTC

Legal pad with scribbles sold for 16 BTC

Go read this piece if you haven’t yet 👇🏼