antfx (@antefex_moon) 's Twitter Profile
antfx

@antefex_moon

VP Growth @hypepartners Marketer | Strategist | Advisor | Bulliever

ID: 1394066824732479488

calendar_today16-05-2021 23:07:02

3,3K Tweet

1,1K Followers

2,2K Following

Route 2 FI (@route2fi) 's Twitter Profile Photo

Oh, so you missed Plasma, Aster, Avnt, Stbl, Hemi, Cards, Pump, Useless, Fartcoin, Grass, WLFI, Trump, GOAT, Virtuals, A16Z, Pnut, Moodeng, SUI, WIF, SPX, Popcat, TAO, Pendle. Just log off and never try again.

n8n.io (@n8n_io) 's Twitter Profile Photo

The Gemini node just got a glow-up! With Nano Banana onboard, you can generate and edit images like never before. To celebrate, we asked Nano Banana AI to go “all out banana” on our UI. Yes, it’s chaotic. Yes, it’s yellow. Yes, it’s probably a terrible idea. But it was too fun

The Gemini node just got a glow-up! With Nano Banana onboard, you can generate and edit images like never before.

To celebrate, we asked Nano Banana AI to go “all out banana” on our UI. Yes, it’s chaotic. Yes, it’s yellow. Yes, it’s probably a terrible idea. But it was too fun
Dan Mac (@daniel_mac8) 's Twitter Profile Photo

Anthropic release Sonnet 4.5 this week. Their first model that scales RL post-training. Similar to what OpenAI did with GPT-5 according to Epoch. 👇 > Tops SWE-Bench and Terminal-Bench > 3 hour tasks on METR Long Duration > Price parity with GPT-5 That's my prediction.

Anthropic release Sonnet 4.5 this week.

Their first model that scales RL post-training. 
Similar to what OpenAI did with GPT-5 according to Epoch. 👇

> Tops SWE-Bench and Terminal-Bench
> 3 hour tasks on METR Long Duration
> Price parity with GPT-5

That's my prediction.
TestingCatalog News 🗞 (@testingcatalog) 's Twitter Profile Photo

BREAKING 🚨: Google started rolling out new Nano-Banana-powered Video Overview styles for NotebookLM! - New Styles: Whiteboard, Watercolor, Retro print, Heritage, Paper-craft, Anime - New Format: Brief First sample here 👀

Alex Prompter (@alex_prompter) 's Twitter Profile Photo

RIP prompt engineering ☠️ This new Stanford paper just made it irrelevant with a single technique. It's called Verbalized Sampling and it proves aligned AI models aren't broken we've just been prompting them wrong this whole time. Here's the problem: Post-training alignment

RIP prompt engineering ☠️

This new Stanford paper just made it irrelevant with a single technique.

It's called Verbalized Sampling and it proves aligned AI models aren't broken we've just been prompting them wrong this whole time.

Here's the problem: Post-training alignment
antfx (@antefex_moon) 's Twitter Profile Photo

crypto AI projects might be dead models just learned how to communicate without tokens imagine communicating with others by syncing up your memory instead of using words... that’s what this new paper from Tsinghua just showed, LLMs can skip text entirely using cache-to-cache

crypto AI projects might be dead

models just learned how to communicate without tokens

imagine communicating with others by syncing up your memory instead of using words...

that’s what this new paper from Tsinghua just showed, LLMs can skip text entirely using cache-to-cache
antfx (@antefex_moon) 's Twitter Profile Photo

LLM traffic is shrinking  and that’s the point Everyone’s watching LLM traffic drop and calling it a problem It’s not. It’s the new baseline AI visibility was never about driving clicks It’s about building trust at a scale & making sure your brand, your data, your explanations