2025 Prediction #6: AI Foundation Models Will Become Standard
And why I think we're asking the wrong Containment Question
If foundation models are the most powerful engines yet built, and they’re on their way to becoming ubiquitous in 2025, here’s the real question: Where are we going?
These AI systems aren’t just tools; they’re accelerators. They amplify what we know, what we create, and what’s possible. But if they’re accelerating everything about humanity—our creativity, our curiosity, and, let’s admit it, our flaws—how do we ensure they amplify what’s best about us?
This isn’t just a futuristic thought experiment. By 2025, AI foundation models won’t just be powerful—they’ll be standard. Embedded in businesses, art, culture, and even the devices in your pocket, they’ll shape how we work, live, and create.
What Are Foundation Models?
Let’s start with the fundamentals.
A Foundation Model is a large, pre-trained AI system capable of generating or analysing a wide range of content by identifying patterns in massive datasets. Think of GPT-4 or MidJourney. These systems are generalists by design but can be fine-tuned to specialise in tasks like diagnosing illnesses, writing legal contracts, or composing symphonies.
To put it simply: Foundation models are the Swiss Army knives of AI. They’re versatile, powerful, and designed to adapt.
Now let’s “Explain It Like I’m Five”:
Foundation models are like really smart crayons. Imagine a crayon that can draw anything you ask for—a dragon, a spaceship, or even a picture of your best friend eating spaghetti. And the more you use it, the better it gets at guessing what you want next. That’s what these models do with words, pictures, and ideas.
The Three Domains: Public, Enterprise, and Personal AI
Foundation models thrive across three key domains, shaping our interactions with technology:
Public AI: Everyone’s Tool
Public AI systems are broadly accessible, designed to handle general tasks like answering questions, drafting emails, or creating art.
Example: ChatGPT helping a student write a research paper or someone using DALL·E to create a digital birthday card.
Impact on Culture: Public AI democratises creativity, making tools available to anyone with an internet connection. But it also risks flooding the world with low-effort, algorithmic content—endless AI-generated TikToks, explainer videos, and memes designed for engagement rather than meaning.
Enterprise AI: Tailored for Industries
Enterprise AI takes foundation models and fine-tunes them with proprietary data to serve specific industries. This transforms general-purpose models into specialists for healthcare, finance, media, and beyond.
Example: A fashion brand using AI to predict trends based on customer behaviour, or a hospital using it to generate medical reports.
Impact on Culture: Enterprise AI defines how industries interact with culture. In media, for instance, it accelerates synthetic content production, blurring the line between authenticity and automation. Imagine watching a film where the script, visuals, and even performances are AI-generated—a triumph of technology, but is it art?
Personal AI: Built for You
Personal AI systems tailor foundation models to your individual habits, preferences, and data. Think of Apple Intelligence or a future system that anticipates your needs before you even know them.
Example: A personal AI assistant recommending a playlist based on your mood or rescheduling meetings to reduce burnout.
Impact on Culture: Personal AI isn’t just convenient; it changes the fabric of how we live. It introduces hyper-personalisation but also risks reinforcing echo chambers and reducing serendipity—those moments when something unexpected, and deeply human, enriches your day.
Presonal.AI
Foundation Models: Powering Arts, Culture, and Behaviour
Foundation models are not just reshaping industries—they’re reshaping culture itself. And because culture dictates how we think, act, and connect, their influence runs deep.
Art and Creativity: A New Renaissance or a Karaoke Machine?
Foundation models democratise creativity. A teenager can use MidJourney to create surrealist art that once took years to master. A filmmaker can storyboard an entire movie in hours using text-to-video AI. But what happens when anyone can imitate a master’s style or produce endless variations of the same aesthetic?
Example: An artist spends years developing a unique style, only to see an AI replicate it in minutes. Their work is no longer irreplicable; it’s remixable.
This raises an existential question: Is creativity about originality, or about curation and meaning-making in an oversaturated world?
Cultural Flattening vs. Explosion
On one hand, foundation models risk homogenising culture. Algorithms optimise for engagement, flooding us with viral trends and shallow, AI-generated content. On the other, these tools amplify niche voices. Indigenous communities are using AI to preserve endangered languages, creating a cultural archive that would have been impossible a decade ago.
The paradox: Foundation models can either flatten culture into algorithmic sameness or explode it into infinite diversity. The outcome depends on how we wield them.
Cultural Flattening and the Rise of the Algorithmic Median
Keep reading with a 7-day free trial
Subscribe to Strange Loop to keep reading this post and get 7 days of free access to the full post archives.