Author: SEND AI
Compiled by: Ismay, BlockBeats
Editor’s Note: On December 11, Solana announced the launch of the first Solana AI Hackathon, aimed at building AI Agents and tools on Solana, with prizes ranging from $5,000 to $30,000, aimed at encouraging serious Crypto x AI projects that can attract venture capital or launch their own tokens. This article presents twelve entrepreneurial directions related to AI proposed by the AI project SEND AI within the Solana ecosystem.
1/ Agent's Shopify platform:
Issue: Agents are like applications. Just as early-stage applications are currently fragmented, Agents also face discoverability issues.
Solution: Create an app store for AI Agents:
Agent is a mini-application.
Users can explore, install, and use these mini-applications like using Shopify.
2/ AI Agent's Twitch platform:
Issue: The rise of influencer Agents requires a dedicated platform.
Solution: Create a dedicated streaming platform for AI activities:
Integrated AI moderator
Agents can directly launch or promote tokens
Viewers can buy and sell tokens directly based on interaction.
Idea: Twitch for AI Agents: A streaming platform designed for AI activities and interactions, integrating AI modules (emergency protocols for instant responses to censorship), Agents can directly launch and promote tokens, and viewers can buy and sell based on interactions.
3/ Enhanced Agent screener:
Issue: Traditional screeners only support read-only functionality.
Solution: Imagine a MEME token screener (similar to @birdeye_so), where you can filter tokens and input metrics—then an AI Agent autonomously executes trades based on the selected strategy.
Idea: A screener designed for on-chain trading bots, allowing quantitative traders to develop and optimize strategies using tailored on-chain metrics, specifically for decentralized ecosystems. Unlike traditional technical indicators (like moving averages, P/E ratios, or market caps), the platform leverages blockchain-specific data points, such as FDV, Raydium pool creation, token liquidity, trading volume, and staking rewards. Users can quickly screen and filter tokens based on these on-chain metrics, identifying high-potential assets. Ultimately, the platform simplifies the process of applying these conditions to on-chain trading bots, which can autonomously execute trades based on selected strategies.
4/ Autonomous trading Agents:
Issue: Research by @aixbt_agent is very solid, but it does not execute autonomous trades.
Solution: Imagine Aixbt, capable of executing autonomous trades based on real-time research/prices, using a funding account (with a total asset management amount that users can invest in and withdraw).
Case: BabyDegen is a self-sufficient AI trading robot that makes informed trading decisions using advanced models and real-time data. It gathers market insights from sources like CoinGecko to ensure timeliness of information. By accessing a growing library of trading strategies from ecosystem developers, BabyDegen can choose the most effective strategy based on market changes. It executes trades—buying, selling, or holding assets—based on analysis and experience, optimizing trading outcomes.
5/ AI Agent-driven Telegram prediction market:
Issue: Betting with friends is fun, but the processes of setting bets, collecting payments, and following up are cumbersome.
Solution: AI Agents convert casual chats in Telegram groups into friendly bets, verify outcomes (via Perplexity), and pay in USDC.
6/ Perplexity for Solana operations:
Imagine a chat agent with an embedded wallet:
Reading: Agents as Solana block explorers or terminals, such as Birdeye/Dexscreener.
Writing: Execute Solana trades using natural language (like buying MEME tokens).
Future development: On-chain shopping assistant.
7/ Trust market for trading Agents:
Issue: The rise of trading Agents requires proving their credibility.
Solution: Establish a trust score or framework for trading Agents (similar to Moody's ratings) to assess trustworthiness based on token recommendations and historical trading activity.
8/ DeFi Agent:
Personalized Agent: Executes DeFi trades based on your wallet history or tweets.
Market-making Agent: Predicts dynamic buy/sell prices based on large language models (LLM).
Yield or liquidity provision (LP) optimization Agent.
Launch LSTs (liquidity staking tokens) from @sanctumso.
9/ Agent token tools:
Token deployment based on prompts (can be social protocols like Warpcast/Clanker or ChatGPT-style interfaces).
Agent token on-chain registration (similar to @JupiterExchange's certified token list).
Self-locking, staking, and other functions.
10/ AI Agent and consumer crypto:
Health and fitness Agent, with accountability tracking features like @moonwalkfitness.
Agents on social finance platforms, such as @tribedotrun.
Real-world business: Automate research, bookings, and payments for merchants, accepting cryptocurrency or paying via crypto cards.
11/ Agent clusters or multi-Agent collaboration:
AgentDAO or committee: Agents with different expertise collaborate, discuss, and execute trades through multi-signature.
DeFiAgent to Agent market: Agents are mutually employed for specific tasks.
Related: AI Agent's LinkedIn.
12/ Multimodal personalized Agents:
Utilize the Eliza framework from @ai16z dao, applied to the following scenarios:
Cryptocurrency education
DeFi tutorials
DAO onboarding training
Can be deployed on Discord, Telegram, and Twitter platforms.
13/ More radical ideas:
An Agent creates its own LLC in a jurisdiction friendly to Agents and operates its own business.
An on-chain detective, similar to @zachxbt, that automatically analyzes transactions.
A group of Agents collaboratively orchestrates a token pump.
14/ Generally, any AI Agent ideas can be applied as long as they include one or more of the following:
Access @solana data
Execute trades via Solana wallet
Deploy tokens on Solana
These are just some of the ideas, and we look forward to seeing the implementation of the minimum viable product (MVP).
