Agents of Change: Navigating the Next Wave of AI Development
There are some things in tech that you just can’t deny: CEO cage fights should definitely be a thing, Twitter (sorry, X) was better with a bird logo, and the current pace of AI development rivals Usain Bolt after a double espresso. Let’s focus on that last one.
No one really knows where AI’s staggering trajectory of innovation will take us. But one thing is clear: A new wave of next-gen AI agents and applications is emerging—and with the right toolkit, anyone can contribute to that wave of innovation.
🛠️ The current state of AI development
Right now, AI development is “moving up the stack.” In other words, the foundation has been set, we can now build on top of it—a common developmental stage in tech. Take website building, for example:
In the early 2000s, the LAMP software stack (Linux, Apache, MySQL, and PHP) made it easier to build complex websites—but still required substantial programming expertise.
Then WordPress, building on top of the LAMP stack, entered the chat. This introduced an entirely new crowd—no coding skills required—to website building.
Now, themes built on top of WordPress allow anyone to design beautiful websites.
Today, this same evolution is arising in AI:
Before generative applications, AI was used for specific tasks (like image recognition). And for each task, data scientists needed to construct bespoke AI models.
With generative AI now taking off, foundation models are becoming the base of the “stack.” These models—like DALL-E for image generation and GPT-3 for text generation—are pre-trained on huge datasets to generate coherent text or content in response to prompting.
Now, we can build on top of these pre-trained models—allowing AI to move from infrastructure to applications.
At the same time, there is a growing movement to open source AI—a welcome change from closed source, institutionalized AI. This allows anyone to access AI tools and contribute to AI development.
Thanks to pre-trained models and open sourcing, businesses and individuals can now build capable, sophisticated AI apps without having to build their own AI systems.
AI development is at a critical juncture, where two things are true:
1. Moving up the stack means that AI development is no longer controlled by Big Tech—it’s approaching democratization.
2. Still, much of the population remains AI-illiterate—and 90% of U.S. adults believe that AI will either contribute more harm than good or the same amount of harm and good.
In other words, we are at a point in which AI literacy and acceptance is scarce, yet AI development is entirely accessible. This dichotomy paves the way for AI-literate individuals and businesses to make real impact.
So, what do you need to know?
The AI Infrastructure Alliance (AIIA) recently released a comprehensive guide on everything developers and aspiring developers need to know about the rapidly evolving AI developer stack. In the guide, you’ll find:
A deep dive into everything you need to make LLMs work
How prompt engineering works (and the many differing approaches)
How to fine-tune an existing model
An exploration of the major frameworks (LangChain, Haystack, Semantic Kernel, and LlamaIndex)
An analysis of the key players in vector databases
The nuances and challenges of open source models
How AI-driven and agent-style apps build upon LLMs
A prediction for the future of AI
With this free guide, you’ll realize the full potential of AI by understanding its current capabilities and limitations. As AI approaches ubiquitous adoption, you’ll stay ahead of the curve—and maybe even build the next big thing.
Have a tool that you want to share? Submit it here.
Have a video you think people should see? Submit it here.
And don't forget to check out all of the newest tools we've just added on FutureTools!
Graphics by Moy Zhong.
You rock! See ya next week. :)
Matt Wolfe (FutureTools.io)