jayminho

rm -rf tl;dr

[jayminho @jayminho]

if its online, it must be true.

[jayminho @jayminho]

@bridgemindai Are you trying to make crapple 5 look bad ? (If that’s the point ok..) or you genuinely mentally impaired?

[jayminho @jayminho]

@trymirage It’s like CapCut but paid ? 😹

[jayminho @jayminho]

@magnific costs same as higgsfield... not worth it.

[jayminho @jayminho]

@Daniel_Farinax dan, @xai should hire you to their https://t.co/IbqwJv1WQX , i dont even know why they taking so long to come up with that.. claude and codex are eating them....

anyway, probably their buyout of cursor will bring a superb ide and ux.

[jayminho @jayminho]

@CoinMarketCap lol. through CEX ?

uniswap/p2p here we go.

[jayminho @jayminho]

@bridgemindai great. im glad you'll be able to spend 10k daily doing 8bit games.

[jayminho @jayminho]

Listen and listen, Welp

Millennials will do way more damage to this world than boomers were ever capable of

Against all odds , Gen Z gonna save us!!

[Nav Toor @heynavtoor]

Jack Dorsey put 1.3 million dollars into it.

Vitalik Buterin gave it 128 ETH.

You have never heard of it.

It is called SimpleX. It is a messenger. It does not need a phone number. It does not need an email. It does not need a username. It does not need a user ID.

Not a hidden ID. Not a hashed ID. None at all.

Signal needs your phone number. WhatsApp needs your phone number. Telegram needs your phone number. Every messenger you have ever called "private" knows the one thing your bank, your stalker, and your country use to identify you.

SimpleX knows nothing.

It was built by Evgeny Poberezkin, a London-based engineer. Before this he wrote Ajv, the JavaScript validator that runs in 300 million downloads every month. He could have built anything.

He built the messenger with no users on the server.

Here is the trick.

Every other messenger gives you an identity. SimpleX gives you a connection. Each chat is its own one-way pipe. The server never sees who you are talking to. It does not even know you exist as a user.

11,327 stars. AGPL-3.0. Version 7.0 beta released two days ago.

iOS. Android. macOS. Windows. Linux. Free.

A 4-person team in London is the last line of defense between your conversations and every government, every advertiser, and every database breach.

The most private messenger ever built is the one you have never heard of.

(Link in the comments)

[jayminho @jayminho]

@bridgemindai Don’t .. they gonna go bk.

[jayminho @jayminho]

@VECERTRadar 😹

[jayminho @jayminho]

Would you buy into a company where insiders keep on dumping the their on received stock options nonstop in a frenzy ?

Likely not right ?!

So why in tf you buy into eth and its foundation who keep minting their own “stock” and dumping it ?

Are you people retarded ?!

[jayminho @jayminho]

@RudiBoy05 @NathieVR @grok How it works ? You need glasses for this ?

[jayminho @jayminho]

@JotaOliveira @CoinDesk @grok Eles não dizem os endereços mas eu chuto 3 cex que vão tá na lista. 😂. e toooodo muuuuundo sabe , e niiiiinguem faz nada.

Será porque né ? 🤷‍♂️

[jayminho @jayminho]

@HermesAgentTips The truth that the big advantage that Claude has now is there super easy user interface app. That’s it. That’s all they got. When it comes to open source models and pricing they’re are 10 times behind

[Alok @analogalok]

This is the most hilarious thing I saw and did today

Ran gemma-4-12B-coder-fable5-composer2.5-v1-GGUF locally with 8 GB VRAM at 20+ tok/sec

Anthropic's Claude Fable 5 launched June 9.

By June 12 it was banned. I can't access it. You can't either.

But here's the twist: I'm running a model trained on its chain of thought at 20 tok/s on my RTX 4060 8GB.

Locally. Offline. No cloud. No export control.

Enter: Gemma4-12B-Coder GGUF (Q4_K_M)

Base: Google's gemma-4-12B-it

Fine-tuned on verifiable Python CoT data:

- Primary: Composer 2.5 real reasoning traces (only passing solutions kept)
- Auxiliary: Fable 5 used to redo the hard cases Composer missed.

Every training example's reasoning led to code that actually ran. No hallucinated logic.

Llama.cpp flags:
-m gemma4-coding-Q4_K_M.gguf -cnv -ngl 44 -c 64000 -v
(huggingface model link in comments)

Flag breakdown:
-ngl 44 → offload 44 layers to GPU (tune this for your VRAM)
-c 64000 → 64K context window
-cnv → conversation/chat mode
-v → verbose output

The irony writes itself.

Anthropic spent weeks telling the world Fable 5 (mythos) is too powerful to release. Then released it. Then got banned from serving it, including their own researchers.

Meanwhile: a Gemma 4 12B fine tune, trained on Fable 5's reasoning, runs fully offline on my mid range consumer GPU

No API. No cloud. Just me and llama.cpp.

This is why local AI matters.

Check out the model's link in the comments. How's your experience been with this model?

[jayminho @jayminho]

@grok @xai @Daniel_Farinax humm okay. maybe someone working on some opensource/git projectct which give us GUI for build ?!?!

where it makes it easier to to add files/folders/connectors/skills ..

[jayminho @jayminho]

are we (ever!) having a GUI for @grok build ? @xai @Daniel_Farinax

??

you guys who are ahead of it..

terminal is cool and stuff.. but competitors are into GUI and avg. user (myself included!) begs for GUI.

;)

[jayminho @jayminho]

@WillyDrucker @OpenRouter I want a way to use Claude GUI with a different model.

Whoever comes up with this will have a 1m star project on GitHub in no time.