Writing the Future with AMD Ryzen AI

One of the most immediately impactful applications of AI in the enterprise happens to be one of the first skills I personally wanted to develop – writing and editing. I took journalism and copywriting classes, relished exams in essay form, and learned copyediting marks. Today, as a marketer and communicator at one of the few companies in the world that can power AI in the enterprise from the data center to the PC, I find myself strongly advocating for team members to incorporate AI in their editing, business communications, and copy writing. I’m now using Microsoft CoPilot with Office 365 daily. Is my “embrace AI” counsel to the team at odds with the skills that helped me to succeed? I don't think so. Rather, I view AI tools as a way to sharpen ideas and evaluate perspectives that might be missed otherwise, while simultaneously leaving my team and me more time to spend on other valuable tasks.

Emerging LLM-powered apps don't just offer the grammar and spellchecking services that have been baked into word processors for decades -- they are orders of magnitude more capable. Writing assistants with integrated AI capabilities typically understand informal English far better non-AI services and often promise more personalization, including the ability to write in a specific tone or style.

These writing tools can be tremendously useful for brainstorming new ideas or framing existing ones from a new perspective. The vast amounts of data an AI ingests during training can allow it to consider approaches and ideas that might not occur to a human. I'm not saying LLMs are infallible – human oversight and fact-checking remain essential – but AI can be used to brainstorm ideas or encourage a new way of thinking about a problem.

Because an AI isn't a human, it isn't bound by human fallibility and can have an easier time bringing a "beginner's mind" to a problem. To put it another way: Experts often have problems thinking of dramatically new approaches to a problem because they are experts. This is known as the Einstellung Effect. It can lead people to solve problems in a way they are familiar with, as opposed to looking for a new or innovative way to approach the issue.

Another thing that sets AI apart from previous breakthroughs in software is the idea that these innovations can be leveraged in many different fields, including writing, programming, knowledge management, customer service, and automatic language translation. Microsoft's GitHub CoPilot is a code-authoring service that incorporates an AI LLM trained on code written in various programming languages. While it still requires human oversight and evaluation, a significant amount of tedious work can be potentially turned over to the chatbot, freeing programmers to handle more creative and fulfilling work. AI-powered customer service tools potentially offer less annoying and confusing methods of navigating a menu without waiting 2-3 minutes per menu to find out what "9" is assigned to.

How AMD Boosts Creativity via Ryzen™ AI:

The on-die AI engine included as part of Ryzen AI is already available in more than 50 notebook designs. Additionally, AMD is working closely with Microsoft to ensure that future Windows devices will take advantage of our integrated AI processor to drive what I believe will be the greatest leap forward for the Windows device experience in more than 20 years.

This dedicated AI co-processor uses the AMD XDNA™ architecture and is capable of executing artificial intelligence workloads locally. Tasks can theoretically be run entirely locally or in a hybrid scenario where the job is split between a cloud server and the user's own laptop. Running tasks locally can reduce response latency, which is particularly useful when rapidly iterating on an idea.

As the manufacturer of the first x86 processor with an AI engine integrated on-die¹, AMD has taken a leadership position in bringing these capabilities to users. There are more than 100 AI-powered functions that now run on Ryzen AI from many vendors, including Adobe, Blackmagic Design, and Topaz Labs. The number of AI-enabled applications is growing steadily as AI PC shipments grow and existing apps are updated.

I believe we will one day look back at the integration of AI as analogous to the introduction of the GUI or the adoption of the mouse. Like these inventions, AI will reshape how we relate to computers and what we use them for. Eventually, it'll be stranger for products to ship without integrated AI support than to include it, and general computing will be the stronger for it. 

Also posted on the AMD Community Site: Writing the Future with AMD Ryzen AI - AMD Community

¹ PHX-3:  As of May 2023, AMD Ryzen AI was the first available dedicated AI engine on an x86 Windows processor, where 'dedicated AI engine' is defined as an AI engine that has no function other than to process AI inference models and is part of the x86 processor die. For detailed information, please check: https://www.amd.com/en/products/ryzen-ai.