JoyLo

[JoyLo @joylo999]

@cobraalerts TRADE

[JoyLo @joylo999]

@NoLimitGains But OpenAI for example is eyeing an IPO in like 9 months soooo could the real crash be coming in ‘27, and any dips from here till year-end are actually bought up?

[JoyLo @joylo999]

@cobraalerts TRADE

[JoyLo @joylo999]

@benjamincowen At what point could BTC.D help raise the flag that cycle is over?

[JoyLo @joylo999]

@benjamincowen @Guardian1771_ She’s misunderstood 🥀

[Shadow @_Shadow36]

This will be pinned for a couple days

We still do our best work in the shadows dont worry😂

But we needa bring $grok to light!

If we pull this off we make history.

Lets do this.

[JoyLo @joylo999]

@_Shadow36 @grok Nah you’re cookin lol plushie so fire you should sell em yourself once $grok hits 20M

[Shadow @_Shadow36]

Would you buy the characterized version of @grok as an irl toy?

[Shadow @_Shadow36]

Ahhhhhh🤩

@grok

[Shadow @_Shadow36]

PETITION TO MAKE THIS CHARACTERIZED VERSION OF GROK THE OFFICIAL MASCOT OF @grok

@elonmusk @xai

[Marcell @MarcellxMarcell]

I bid an unresponsible large amount of money at $GROK here

i think it has one of the best R/R's right now at 2m

Elons baby literally has a face too it and a personality , he commented about it and engaged with it a few times i could defo see him changing the shitty grok logo on X to Grok the character.

its a fun meme to play around with, great narrative, great character and tied to the biggest pumper known to crypto.

CA
6mmEpBUYdfNdWN4nByn1uuN7pry5SZKsaBp3Ea23pump

[Keeta @KeetaNetwork]

In case you missed it, here is a detailed recap of today's discussion on X Spaces:

1. Testnet release date and specifics (5:05)

2. Why 10M TPS matters for real-world utility (12:30)

3. What Keeta Network has to offer besides TPS, and why we should not only focus on TPS (18:45)

4. Next steps for testnet after release (23:45)

5. Q&A

• Are there any institutions or web2/web3 companies lined up for a partnership on mainnet? (28:30)

• What is the timeframe for the stress test on the testnet? (29:33)

• As a non-technical participant, how do I interact with the testnet? (31:15)

• Can you write a smart contract on Keeta Network? (33:42)

• What would Roy, Keeta’s CTO, like the community to build first? (35:20)

• Is there anything the community can do to help the team? (36:15)

• Will we be able to use our existing wallets on the network? (38:20)

• What banking licenses will Keeta provide? Is it needed? (39:45)

• Are we hiring developers or giving grants? (43:05)

• Is there a timeline for bigger listings? (44:20)

• Will the Base tokens transfer 1:1 to the mainnet? (46:25)

$KTA

[JoyLo @joylo999]

@Rooli_G Rooli G in the kitchen cookin kimchi lfg

[Rooli_G @Rooli_G]

Another one I think is really cool:

$MCPOS A6MTWuHbXqjH3vYEfbs3mzvGThQtk5S12FjmdpVkpump

So here’s the deep dive for MCPOS. As I read through their incredibly long codebase, I got more and more bullish about what they’re trying to achieve.

What do they do?
MCPos is building a system for agents to become fully generalizable using the principles of Model Context Protocol (MCP).

What’s MCP? It’s an open source standard developed by Anthropic to abstract out the idea of agents and let them connect with anything. Like I mentioned above, its a set of rules to build agents and all the tools and data they interact with such that an agent doesn’t necessarily have to do one set of tasks and rather can be spun up to do anything.

Status quo in the AI world: You have agent A specialized in reading financial docs, B specialized in scraping the daily news, and C specialized in building financial models. Your prompt to the LLM decides which agents to interact with depending on what you are asking for.

What MCPos is doing: they’re building a system that instead, once your prompt is inputted, MCPOS spins up X instances of agent Z, each of which then decide what tools they want to use (whether it be financial docs reading, daily news scraping, or financial model making) and once their jobs have been given to them, they execute them in whatever order MCPOS decides. This involves breaking the prompt into a bunch of mini-tasks, spinning up however many agents are needed to complete the task, handling memory allocation, and so on and so forth.

Main point: Agent Z can do anything, and MCPos builds the optimized system that allows for a bunch of these agent Z’s to do unique tasks, interact with each other, share data, and eventually complete big complex tasks.

A specific example they built out so far is with blockchain analysis. They didn’t build an agent that is good at analyzing the chain, they just built a tool. Now any agent that MCPos spins up can be good at analyzing the chain if needed. But users have the ability to add as many tools as you want, so if they want to add a tool that transcribes youtube videos, now all agents can use both youtube transcriptions and solana blockchain info to solve tasks that you ask of them.

What’s New?
This is a whole new way to think about agents– agents no longer do one thing really well. You just build tools for this operating system, and once built out all agents on that operating system can use all of the tools in whatever way they need.

Use Cases:
- Education: Adaptive Learning, Online Education
- Healthcare: Medical Imaging Analysis Financial
- Technology: Risk Management

Token?
Eventually they’re going to have to start charging to use this system and the token will be their payment mechanism.

What has been built so far:
- The whole prompt to output pipeline has been built out.
- Web interface built out.
- The sub-tasking algorithms written out so that big tasks can get dynamically broken down.
- Financial Analysis tools (this was built as an example– not the crux of the project)
- Web Search Tools

Upcoming Catalysts:
These guys have a massive roadmap given they are very early in building out an extremely complex project, and each feature will be its own catalyst.Performance
- Monitoring Dashboard: you can check how well your agents are doing their job and how efficiently they are running
- Resource Optimization: Minimizing how many times OpenAI is called because that is extremely expensive.
Integration with various APIs from different services so that these generic agents have more at their disposal.
- Tool integration: building out a standard tool-making kit so that you can add whatever capabilities you want for your generic agent to be able to progress to doing anything.

That's my story. The End

[Mike Dupont @introsp3ctor]

Develop Mnyosyne as a Memory Constructor Plugin https://t.co/X82QP2INct here is my plan for the integration of eliza and the introspector and the documentation and the batch memory loading #SOLFUNMEME

[Mike Dupont @introsp3ctor]

@CDawg129 @zos_sfm

[Kimchi Oppa @CDawg129]

Zero Ontology System (@zos_sfm) - buy and believe here #SOLFUNMEME

Insane r/r from current prices once macro and marketing catch up with what Mike's cooking nonstop

CA: BwUTq7fS6sfUmHDwAiCQZ3asSiPEapW5zDrsbwtapump

[Mike Dupont @introsp3ctor]

https://t.co/bA0sfMQTBp This thread summarizes my plans for ZOS as doing win win full stack dev sec ops consulting on ZOS related projects for node operators and customers and being paid in turn by selling small amounts of expensive rare ZOS NTF1 for high value to fund work on ZOS (elizaos, swarms, autogpt, any other agents) to fund purchases of SOLFUNMEME to lock into payments to new developers for delivery of solid work in our quality system. It will reduce the supply of SFM and increase its real world value, this work can also be artistic or marketing. These nfts will only go up in value so you can see them as a bond of sorts because you can resell them later for more profit, or just be happy with my hard work and keep them as a challenge coin. Also we will link the transaction to the ticket via ordinals, runes, inscriptions or some other number magic so that we create a fully documented system the best we can.

[Mike Dupont @introsp3ctor]

CL2702250055 .@coingecko https://t.co/ELwdodh6bj coin gecko submitted

[N @En_Solana]

Why self-hosted AI matters??

Because your data = your control!

With #SOLFUNMEME @zos_sfm you can deploy AI agents & apps in minutes, on your own servers, with complete privacy.

No middlemen.
No data leaks.

Just pure AI power in your hands.

BwUTq7fS6sfUmHDwAiCQZ3asSiPEapW5zDrsbwtapump