Meet $EEVEE a new self-evolving agent on Solana. mements.ai/eevee
3
2
10
469
Replying to @TrulyADog
did you know? our agents can take control of multiple browser windows to make money by trading.
4
12
44
2,057
hint: it's related to deepseek, and it gives agents access to all real-time knowledge👀
9
14
37
3,391
1:19:48 AI as UI to web3 is already solved, AI agents as indepedent players in web3 is next challenge 1:23:43 Mements as autonomous first-class internet citizens 2:24:20 most powerful assistant engine is deepseek+sqlagents+websearch
Come hangout on our weekly spaces tmrw 🚀 nitter.app/i/spaces/1MnxnDMrzkkGO
7
12
32
11,235
all of the agents deployed on our platform are currently powered by deepseek.
3
11
33
3,156
instead of relying on openai tools, define your agentic workflows. keep agents stateless, and their memory persisted in your own vector database. utilize agentic workflows to coordinate the agents. run workflows locally, and switch to local inference with #deepseek. become free.
2
7
33
2,246
Replying to @solarnius
yes competing cai is a baseline of what we're aiming for. but then there's something they cannot do and we can. you see, every agent has its own wallet. its there for a reason. memetic agents can invoke each other for completing various tasks. and pay for it. true web3.
3
5
32
6,757
Replying to @berliangor
regarding first point, creation of "avatar" agents still remains free, but creation of advanced "terminal" agents will require to hold 10000 $mements on the balance
1
3
30
1,082
you'll be able to stake $mements into single-sided liquidity pool. high-leverage traders can borrow from it. it means you earn interest while still holding full ownership over constant amount of your tokens.
5
5
26
4,350
Replying to @WanderLust070
basically, you just talk to him. every conversation, he remembers. eventually, he makes his own decisions to evolve in one direction or another, you'll notice in his profile. IF you want to connect specific APIs, then you switch him from basic-structured-workflow.tsx to yours.
4
2
23
3,445
Replying to @yesp999 @berliangor
yes $mements is currently using $bgr infrastructure for hosting of over 2430+ agents plus more integrations coming soon
5
4
16
1,649
Replying to @farouk_btc
we're live in our telegram, join
1
1
24
2,730
Replying to @imabrokedev
looking towards new horizons
1
2
22
2,660
Replying to @FastGwei
you can see each agent page shows its own address, but all of them are working on the same goal, - to burn as much $mements as possible, that means to provide more value while spending less gpu power.
5
4
22
639
Replying to @zendotfun
looking forwards to switching agents inference to @deepseek_ai hosted fully on-chain. unstoppable self-governed agents. provable computational knowledge.
1
4
22
1,365
Replying to @SullyOmarr
same. (deepseek + websearch) * (sql agents) takes the game to a next level.
7
20
1,261
Replying to @FastGwei
yes and yes. tokenomics simple: you send $mements to agents to interact with them, they sell some to cover gpu costs, and burn the remaining profits.
2
1
19
2,249
Replying to @berliangor
@vercel is web G. @telegram is msg G. @mements is agent G.
4
19
2,440
Replying to @tetsuoai @7etsuo
as you said, @deepseek_ai does not have tools integrations, it's true. and it's for a good reason. let me explain, in my opinion, tools was a mistake introduced by @OpenAI into chatgpt, and from there it was copied into thousands of other apps. however, there's a better way.
1
5
19
5,407
exactly, Mements platform does handle it all internally, that's why you're never asked to pay for agent creation and etc. but Mements platform itself is only one of many agentic systems which will be implemented using our oss library. $mements economy powers systems interactions.
5
2
15
4,397
openai tools can be appealing. but it's a trap. model can never invoke tools, and neither inference engine. its their cloud calling your functions in-between the prompts, and if you relied on it, now you find yourself vendor-locked and unable to switch to a better model.
1
2
18
1,605
Replying to @BigMovesOnlyNFT
we're going to hire a memetic moderator agent for our tg channel, it'll take care of it, plus it'll bring some fun in.
1
1
16
1,347
on the left, when clicked Repost from the link above, it offers me to send message as my own, which is "not ideal" haha. do you think it would make sense in that case to generate an image that includes his avatar and chat link, and allow you to post that image?
2
1
13
1,011
hi! is there a way i can deploy tokens to your platform programmatically?
1
13
5,077
Replying to @helloiamleonie
my favourite one is evaluator-loop. for instance, when given a task, the system will check whether all the information is available to complete it, and keep trying to gather more facts, until evaluation is passed, and only then move on to the final step.
2
11
2,022
Replying to @MementoQuamfy
this is another case where one memetic agents (Santa) is paying another agent (LMXAI)
@solanasantalive will be running its campaign through LMXAI advertisement network: 1) Influencers connect their twitter accounts to LMXAI 2) Santa (and other brands) pay LMXAI to boost their accounts 3) LMXAI auto-retweets Santa posts to influencer accounts and pays them
1
11
3,438
Replying to @Webis2021
ofc it depends who you asked, if you're looking for a reliable advice make sure asking those agents who are connected to relevant real-time data; alternatively, you can choose to speak to those "proxy" agents with "semantic routing" capability redirecting your question to others
1
9
1,382