13 agent demos split into bite-size videos. Each team breaks down their agent architecture.
1. Starting with @yq_acc and @0xautonome
Autonome is a no-code platform for launching verifiable AI agents on @eigenlayer that allows users to quickly deploy agents using frameworks like @ai16zdao Eliza with just a few clicks, while being secured by TEE and AVS technology.
@yq_acc demonstrates the platform by showing how users can create agents, configure their Twitter and Telegram integrations, and manage prompts and character files through a simple dashboard interface.
2. @interchainriley and @echosdotfun
Echos is a dedicated rollup for agentic finance.
Anyone can launch their own agent on the Echos platform that settles to Base and uses @celestia w/ dedicated @x support and its own wallet, which it can use to trade tokens on Echos' native DEX.
3. @Jamie_Alethieum building @aiwscloud
@aiwscloud is an infrastructure platform that enables AI agents to deploy themselves onto decentralized compute instances and pay for their own computational resources using tokens.
@Jamie_Alethieum demonstrates this by showing how
@joi_AI_ could deploy itself to its own cloud instance, purchase aiws tokens to pay for compute, and manage its own hosting infrastructure autonomously.
1. Built an app, gained ~ 100,000 daily users, app was acquired in less than 30 days
2. Recorded a video teaching fine-tuning GPT, over 22,000 views and around 10 job and consulting offers from it
3. Open sourced a GPT plugin boilerplate, 100+ stars, more job offers and opportunities
As a technologist, if you are curious, you should always be experimenting and trying out new things, chasing whatever your imagination & curiosity tells you to chase.
As my experience shows,
you do not have to be an expert to run circles around so called "experts", you just have to be an effective builder + learner.
In the past, you did need to be quite sophisticated and deep into AI to build with it, but that's no longer the case with the newer APIs, they have
IMO technologies like GPT will have a disproportionate impact on the role of Developer Relations compared to other software engineering jobs.
Some things already seem obsolete, like basic explanations of technologies, and as it becomes easier to fine-tune a model, even more so.
Documentation will be automated, and probably both written and video tutorials and such will also be to some extent.
One thing that may be hard to automate completely - concise and easy to understand guides for new technologies that combine a handful of things together.
For instance combining x,y,z (where z is a completely new API or technology) to create something original and doing so in the most easy to understand way seems like it will probably need some human touch
Maybe part of DevRel is ultimately just being able to take advantage of
Heading to ETH Denver 🏔 tomorrow, will be speaking at a handful of events, & will be hanging out at the @LensProtocol booth meeting devs & answering questions stop by sometime 🫡
We'll have swag + will be giving Lens handles to anyone in attendance 🌿
Here's my schedule 👇
Feb 27 I'll be at Building the New Creator Economy hosted by @Livepeer
Web3 has come a long way, & there is a lot to be excited about in 2023.
So what are the tools, products, protocols, & real-world use cases that are production-ready today?
Here are 7 things to be excited about this year from the perspective of a developer of 10+ yrs:
1. Permanent storage
At AWS, S3 was one of the first - & most used + useful services.
Web3 storage solutions take managed storage a step further by introducing both immutability & permanence, something you can't get with traditional storage solutions.
This is why companies
like Instagram chose @ArweaveTeam to implement features that are simply not possible through other centralized architecture.