gulik.sol

soul searching sol enthusiast
gulik.hxd.sol
Taiyo, Pharaoh, Sharx, Honey-Land & Photo-Finish holder
retweet junkie & shitposter
former eth miner ...

[gulik.sol @Martin_Gula]

incredible tek
$BTC in 3 days -10k
74k->64k
please somebody save the shitcoin

[0xMarioNawfal @RoundtableSpace]

GOOGLE'S GEMMA 4 12B JUST BECAME THE NEW LOCAL LLM KING FOR ANYONE WITH AN 8GB VRAM GPU

it runs at 32 tok/s with 64k context on a budget RTX card, no API, no cloud, no subscription, and it's actually smaller in file size than the previous best option.

[TheNewPhysics @CharlesMullins2]

🚨 Scientists just built a refrigerator with NO compressor and NO refrigerant gas.

Just electricity.

Using a multilayer ceramic capacitor, researchers created a solid-state cooling system that changes temperature when an electric field is applied.

The result:
• ~3–4.5 K cooling swings
• works across room temperature
• survives >10 MILLION cycles
• no moving parts
• projected 70–90% Carnot efficiency

This is electrocaloric cooling and it may become one of the biggest threats to conventional refrigeration in decades.

Older materials only worked ABOVE room temperature and needed a brutal 42-day annealing process.

This new PST–PMW material:
• cools down to ~230 K
• avoids the expensive anneal
• handles massive electric fields
• maintains strong entropy transitions

The physics is beautiful.

An electric field reorganizes the material’s internal dipole structure, reshaping entropy inside the lattice and producing a real temperature drop.

Not “cold generation.”
Controlled entropy engineering.

If this scales:
• silent refrigerators
• ultra-efficient chip cooling
• vibration-free scientific systems
• wearable thermal control
• next-gen EV cooling

We may be watching refrigeration evolve from mechanical compression…
to programmable matter.

Follow me if you want the future of physics before it hits mainstream.

[Killa @KillaXBT]

What’s funny is that people labeled me a perma bear just because I was bearish for 8 months.

Now that I’ve flipped bullish after $BTC retraced from 126K to 59K, people seem surprised that I’m not like the other accounts on this app.

I’m here to make money, not defend a narrative. When the market changes, I change. I flip when necessary.

I guess some of you thought calling the top was just a fluke or a lucky guess. And when I catch the move to 160K, I’m sure some of you will call that luck too.

[How To AI @HowToAI_]

Google has published a paper that might end the transformer era.

For the last 7 years, every major AI, ChatGPT, Claude, Gemini, has been built on the exact same architecture: The Transformer.

But Transformers have a fatal flaw.

To remember context, they have to process every single word against every other word. It’s called quadratic complexity. As your prompt gets longer, the compute cost explodes.

The alternative is the old-school RNN (Recurrent Neural Network). RNNs are incredibly cheap and fast, but they have a fixed memory size. If you give them a long document, they get amnesia.

Until today.

Google researchers published Memory Caching: RNNs with Growing Memory.

And it fixes the biggest bottleneck in AI.

Instead of an RNN having a fixed, rigid memory that constantly overwrites itself, Google gave it a "save" button.

The technique allows the RNN to cache checkpoints of its hidden states as it reads.

The memory capacity of the RNN can now dynamically grow as the sequence gets longer.

They built four different variants, including sparse selective mechanisms where the AI actively chooses exactly which checkpoints matter most.

The results rewrite the rules of efficiency.

On long-context understanding and recall-intensive tasks, these new Memory-Cached RNNs closed the gap with Transformers.

They achieved competitive accuracy without the explosive, quadratic compute cost. It perfectly bridges the gap between the cheap efficiency of an RNN and the massive capability of a Transformer.

We have spent billions scaling Transformers because we thought they were the only way an AI could remember a long conversation.

But Google just proved we don't need to process the whole history every single time.

We just needed a smarter cache.

[rari @0xwhrrari]

Google's former CEO just said what everyone in AI already knows

Building wealth is getting easier if you actually learn the tools

Not by scrolling AI threads

By understanding agents, Claude Code, prompts, memory, skills, MCP, and routines

Save this before it disappears from your feed

Everything below is free:

Agent architecture
https://t.co/hhrSyF3y6u

Claude Code 101
https://t.co/71Zu7xQmxc

Claude Code in action
https://t.co/vg8fPChJxl

Prompt engineering
https://t.co/xEu7M3CPe1

Interactive prompt course
https://t.co/oGCxHkqFBD

Claude.md + memory
https://t.co/v6JKlS4OJl

Skills
https://t.co/AJCWlDcMU7

MCP
https://t.co/tQTUYCwN16

Routines
https://t.co/gAnEdq5SP0

Claude Code ultimate guide
https://t.co/WeDLJBVaGM

Awesome Claude Code
https://t.co/wvpwfcaf6U

Anthropic academy
https://t.co/uu3JyZyWyt

Official Claude Code docs
https://t.co/xuUWAfZGXy

All of this costs $0

Most people will keep asking AI one question at a time

Save this and start learning the stack

[Flying Dutchman @DeadSeaTrader]

2-month scenario completed 😈

Now the question is simple

Do you guys want the next $BTC plan?

[Leonard Rodman @RodmanAi]

SAM ALTMAN HAS A NEW PROBLEM. 🤯

Google just shrunk 31GB of AI memory down to 4GB.

The tool is called TurboVec.

It uses up to 16x less memory, searches faster than FAISS, runs fully offline, and works on a regular Mac.

No expensive GPU cluster.
No cloud dependency.
No compromise on speed.

→ 16x lower memory usage
→ Faster vector search
→ Works with LangChain & LlamaIndex
→ 100% open source

The race to build bigger AI models is loud.

The race to make them dramatically cheaper just got a lot more interesting.

Repo: https://t.co/08TFGtHL6K

[Lookonchain @lookonchain]

Arthur Hayes(@CryptoHayes) called $ZEC, $NEAR, and $WLD.

He sold near the top, then disclosed his exit and turned bearish.

$ZEC, $NEAR, and $WLD are now back to where they were before his calls.

[Binance @binance]

First, you downloaded Binance on Apple.

Now, you can buy Apple on Binance.

[Malaton @MalatonKnows]

🚨 All $SERIOUS holders please join the tg and the group chat on X ! Have some news to announce soon

$SERIOUS Portal: https://t.co/xA02nQs2Tc

[Killa @KillaXBT]

I've been waiting patiently for this moment... $BTC

The final phase. The final extension.

I'm about 90% invested.

[BONK.fun @bonkfun]

strapped in for the biggest green candle humanity has ever witnessed

[Elon Musk @elonmusk]

Join our AI recommendations & advertising engineering team!

[Wrenn 🧲 🐕 @Wrenndotsol]

63 whales who officially have a minimum of .2% (2 million tokens)... at a MINIMUM!!!

you can do the math how we got this on lock

[Arthur Hayes @CryptoHayes]

This chart is going in the wrong direction. Dumped $WLD. I’m out. See y’all at the clerb.

[Malaton @MalatonKnows]

3k solana:H6qnGp5anYgMJYAaPXKSqKVYa6mKDrn1ruAeK5Dmbonk holders. Thank you. Get ready for next steps. Higher 🚀

[Binance @binance]

A 10M USDC reason to start your investing era.

Complete your first U.S. Stock or ETF transaction and be eligible for rewards of up to 10M USDC.

Read more here 👉 https://t.co/cyRffjBt4L

[Nom @TheOnlyNom]

The hardest thing in crypto isn't launching. It's giving people a reason to care 3 years later.

[Binance @binance]

You buy the stock.
You own the stock.
You receive the dividend.

Just as it should be.

The Financial Super App allegations were true. 🫡

[DEXchange @DEXchangeNOW]

Just heard about the $ZEC exploit

If you want privacy just use us. Dont buy a coin for thesis of privacy. Use the tools.

Buy $KORI who built us if you want to buy the coin ;)