Prompt Protect

[Sam Altman @sama]

it's not enough to just fix the bad things; we need to aspire to a future that involves building new amazing things

a lot of tech (and public policy) has forgotten this

[Chris Frantz @frantzfries]

This is the craziest tweet on this site.

OpenAI engineer calculates the cost to build a Google killer and comes to $50m for the embedding.

$50m for a ChatGPT that knows everything that’s on the internet, almost instantly.

Unbelievable.

[Paul Buchheit @paultoo]

Google may be only a year or two away from total disruption. AI will eliminate the Search Engine Result Page, which is where they make most of their money.

Even if they catch up on AI, they can't fully deploy it without destroying the most valuable part of their business!

[c0ffe3caf3 @c0ffe3caf3]

@virtualharby Fan Horror (upgraded CPU fan @namrednels_gl) tried to do surgery on GPU fans but it went wrong, case fan seems to do a better job anyway. Out of frame: 8 pin PCIE GPU wire connected to 6 pin + a bodge wire... PSU also need upgrading it's like 300W

[Kevin Schawinski @kevinschawinski]

Software ate the world.

Now AI is eating software.

[Slava Bobrov @slava__bobrov]

Two brain cells having a chat:

#neuroscience

[Michael Royzen @MichaelRoyzen]

@AuthenticExit We never use the GPT-3 API for user queries. Full stop. Every search request served on https://t.co/aW7TuRuM8v uses our own models.

[Leo Maxime Brunswic 🇨🇵🇪🇺🇨🇦🇨🇳 @LeoBrunswic]

@mathladyhazel I turned this into an exercise for some oral examations. That's a great illustration of the difference between convergence in C0 and C1.

[The Cybersmile Foundation @CybersmileHQ]

A note to self…

[OpenAI @OpenAI]

We're publishing a report, co-authored with @CSETGeorgetown and @stanfordio, analyzing how language models could be misused for disinformation and how to mitigate those risks. https://t.co/nHiVp7GoxI

[Javi Lopez ⛩️ @javilopen]

I've been rendering in MidJourney for over 200 hours but I didn't find out about this trick until yesterday!

MidJourney can receive not just one but SEVERAL input images and it will come up with a new one mixing up the "high level concepts" 🤯🤯🤯

Here's how!👇

[Andrej Karpathy @karpathy]

Didn't tweet nanoGPT yet (quietly getting it to good shape) but it's trending on HN so here it is :) :
https://t.co/qouvC6xuXq
Aspires to be simplest, fastest repo for training/finetuning medium-sized GPTs. So far confirmed it reproduced GPT-2 (124M). 2 simple files of ~300 lines

[Jim Fan @DrJimFan]

Many people don’t understand how challenging Minecraft is for AI agents.

Let me put it this way. AlphaGo solves a board game with only 1 task, countably many states, and full observability.

Minecraft has infinite tasks, infinite gameplay, and tons of hidden world knowledge. 🧵

[Google DeepMind @GoogleDeepMind]

Introducing DreamerV3: the first general algorithm to collect diamonds in Minecraft from scratch - solving an important challenge in AI. 💎

It learns to master many domains without tuning, making reinforcement learning broadly applicable.

Find out more: https://t.co/7GP0R93Yvc

[Pete @nonmayorpete]

Update on OpenAI/Microsoft deal terms:

Once $92 billion in profit plus $13 billion in initial investment are repaid [to Microsoft and once the other venture investors earn $150 billion, all of the equity reverts back to OpenAI

[Alex Kantrowitz @Kantrowitz]

ChatGPT is getting training updates and lying about them

[Meta for Developers @MetaforDevs]

Read about the new tools and steps we’ve taken at @Meta to reduce the time waiting for code reviews, which increased overall engineer satisfaction and productivity without sacrificing review quality. https://t.co/hK8poalZfr