A Positive Side of AI
Artificial Intelligence is here to stay so how can we find ways to live with it?
As someone watching the evolution of AI closely, it's been fascinating. It's also been pretty scary. It feels like things could get worse just as easily as things could get better. ChatGPT, Lensa, and Midjourney show us that an AI future is right around the corner. Recently, I came across a few ideas and concepts that make me feel much better about the future of AI and how we, as society, might adapt to the widespread availability and adoption of AI.
One of the things that makes AI so scary is that it's hard to know what it's capable of today. On Stable Discussion, we're going to spend a lot of time on this topic and help map the boundary of possibility. Looking at the new AI products (ChatGPT, Lensa, Midjourney) the clearest capability of AIs is the ability to create something new based on text or images.
These "Generative" AIs can do many different things that are extremely compelling:
Write essays and synthesize based on a number of sources
Create images in your likeness or following the styles of other artists
Create music in many different styles
Create videos and follow direction from the poses in another video
Project the faces of your favorite actors onto the characters in a movie
They can fill out forms on your behalf following your guidance
Not only are they capable of this huge degree of creative potential, but they can also take your feedback on their output to improve their generation!
Once you start looking through this list, the negatives can easily bubble up. Will students use this to cheat on their papers? Will this destroy any potential future for artists? How many jobs will be lost because of this technology?
There are a lot of good questions there. There's a sense of hesitation too and a desire to stop things until we better understand what is going to happen. But these changes are happening, and they're here to stay. Much of the research is open source and widely available. Efforts to restrict or contain AI growth will drive innovation to more open-minded nations or underground, where these capabilities might be misused against anyone unprepared for these changes.
AIs aren't perfect, which makes them even a little bit more chaotic. A Stratechery article titled AI Homework describes how AI's can become confused about subjects, and ChatGPT can reply with incorrect and misleading answers. These "hallucinations" can confuse the unwary, and we need a certain degree of preparedness.
All of this sounds scary, and it's wild to know that we're seeing things today. But that is also relieving. We don't need to wait. We can start to react and prepare based on what we know. There is, and will continue to be, a flood of new information released by AIs.
We cannot prevent this spread, and to a large degree, we cannot protect ourselves from it. In describing this problem, a friend in cybersecurity sent this resource my way. FEMA's national preparedness goal defines steps to prepare for risks with its goal:
“A secure and resilient nation with the capabilities required across the whole community to prevent, protect against, mitigate, respond to, and recover from the threats and hazards that pose the greatest risk.”
As we walk through the steps, it's clear where we are at: Mitigation.
We can mitigate the impact of widespread AI generation. How we do that is by educating ourselves and our families and adapting to these changes in our schools and places of business.
An earlier article by Stratechery, Zero Trust Information, describes how we could transition, as a society, to believe all information is potentially misleading until proven accurate. This puts the power in our hands to control the information we receive. Control we probably should have been training earlier rather than handing off that duty to churches, nations, corporations, news organizations, and even schools.
Schools will look to incorporate AIs into their curriculum. It's a tool students will have in the future, and the best way to prepare them for that future is to integrate this use early on. I believe that some writing will remain in English and Literature curriculums with work needing to be shown, but in many classes, AI could be leveraged so that students can focus on the course details.
AI art has an exciting future, and I think two ways intrigue me. One is the digital signing of art to trace the creation to an original creator. This helps to "prove" the artist's ownership of a piece.
The second is a societal change to value the story of an art piece over just its composition. If we believe that surface-level values are everything to art, then I believe we lose a part of the humanity that created that art in the first place. That's something AI should perhaps take over. The inhuman aspects. Both of these topics deserve their own posts entirely.
While AIs could replace jobs, many jobs may become much easier as previously complex or time-intensive tasks can benefit from an AI assistant. Noahpinion has a great take on this in Generative AI: autocomplete for everything. I specifically like the quote:
This quote is a great way to think about AI's future and how it might affect us. We don't need to fear AIs replacing us but embrace them as tools that can help us with mundane and exhausting tasks. They can also allow us to create and explore ideas in ways we never could before.
AI is here, and it's here to stay. We have an opportunity to shape our future with AI, but we need to be prepared and adaptable. The time for change is now!
 | A guest post by
|