Author: SEND AI
Compiled by: Ismay, BlockBeats
Editor’s note: On December 11, Solana announced the launch of its first Solana AI hackathon, aimed at building AI agents and tools on Solana, with prizes ranging from $5,000 to $30,000, to encourage 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 SEND AI from the Solana ecosystem.
1/ Shopify platform for Agents:
Problem: Agents are like applications. Just like the early stages of applications, Agents currently present a fragmented state, facing discoverability issues.
Solution: Create an app store for AI Agents:
Agents are mini-applications.
Users can explore, install, and use these mini-applications like they would on Shopify.
2/ Twitch platform for AI Agents:
Problem: 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
Audiences can buy and sell tokens directly based on interaction.
Idea: Twitch for AI Agents: a streaming platform dedicated to AI activities and interactions, integrating AI modules (emergency protocols for instant response to censorship), where agents can directly launch and promote tokens, and audiences can trade based on interaction.
3/ Enhanced Agent filter:
Problem: Traditional filters only support read-only functions.
Solution: Imagine a MEME coin filter (similar to @birdeye_so), where you can filter tokens and input metrics—then an AI Agent executes trades autonomously based on the selected strategy.
Idea: A filter designed for on-chain trading bots, allowing quantitative traders to develop and optimize strategies using tailored on-chain metrics, specifically serving decentralized ecosystems. Unlike traditional technical indicators (like moving averages, P/E ratios, or market caps), this platform leverages blockchain-specific data points such as FDV, Raydium pool creation, token liquidity, trading volume, and staking rewards. Users can quickly filter and screen tokens based on these on-chain metrics to identify 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:
Problem: Research by @aixbt_agent is robust, but it does not execute autonomous trades.
Solution: Imagine Aixbt, capable of executing autonomous trades based on real-time research/pricing, using a funding account (with user-investable and withdrawable asset management total).
Case: BabyDegen is an autonomous AI trading bot that makes informed trading decisions using advanced models and real-time data. It gathers market insights from sources like CoinGecko to ensure timely 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 to optimize trading outcomes.
5/ AI Agent-driven Telegram prediction market:
Problem: Betting with friends is fun, but setting up bets, collecting payments, and following up is cumbersome.
Solution: AI Agents turn 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:
Read: Agents as Solana block explorers or terminals, such as Birdeye/Dexscreener.
Write: Execute Solana trades using natural language (e.g., buy MEME coins).
Future development: On-chain shopping assistant.
7/ Trust market for trading Agents:
Problem: The rise of trading Agents requires proof of 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 activities.
8/ DeFi Agent:
Personalized Agents: Execute DeFi trades based on your wallet history or tweets.
Market-making Agents: Predict dynamic buy/sell prices based on large language models (LLM).
Yield or liquidity provision (LP) optimization Agents.
Launch LSTs (liquidity proof tokens) from @sanctumso.
9/ Agent token tools:
Token deployment based on prompts (which could be social protocols like Warpcast/Clanker or ChatGPT-style interfaces).
On-chain registration of Agent tokens (similar to the certification token list of @JupiterExchange).
Self-locking, staking, and other functions.
10/ AI Agents and consumer crypto:
Health and fitness Agents with responsibility tracking features like @moonwalkfitness.
Agents on social finance platforms, such as @tribedotrun.
Real-world business: Automatically research, book, and pay 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 via multi-signature.
DeFiAgent for the Agent market: Agents mutually hire each other for specific tasks.
Related: LinkedIn for AI Agents.
12/ Multimodal personalized Agents:
Utilizing the Eliza framework of @ai16z dao, applicable to the following scenarios:
Cryptocurrency education
DeFi tutorial
DAO onboarding training
Deployable 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 autonomously.
An on-chain detective, similar to @zachxbt, automatically analyzes transactions.
A group of Agents collaboratively manipulates token prices.
14/ Generally, any AI Agent idea can be applied as long as it includes one or more of the following:
Access @solana data
Execute trades through Solana wallet
Deploy tokens on Solana
These are just some ideas, and we look forward to seeing the implementation of minimum viable products (MVPs).
