The AR + AI Race Non-finale | A writers reckoning
A new path towards meaning
Welp, that’s what I get for taking two months off (okay maybe it was three. No real excuse, other than to say life happened in some of the most challenging ways…)
The AR + AI landscape has since shifted dramatically, rendering this essay series obsolete, at least in its original spirit & form. It was going to be a four-part series, predicting the outcomes of the race between Apple and Meta. It took you on a meandering journey up and down the tech stack, analyzing their strength & weaknesses at each layer; from hardware, to software, to AI. It also accounted for leadership, culture, and business philosophy.
Here is Part I and Part II if you’re curious. Each received a solid response from the community and I was eager to crank out the rest… but then a variety of things happened:
1/ Google pulled a Michael Jordan, rocking #45 and bursting back on to the XR scene with AndroidXR and an acquisition of HTC’s XR engineering team. Leaving Google out of the equation suddenly felt silly (although some would argue they never left).
2/ Apple apparently cancelled their AR glasses project (per this Bloomberg report).
Then they apparently didn’t (per a variety of articles like this).
And then I realized who the hell knows. Apple is more secretive than the CIA and any predictions involving their most veiled ambitions are rife with speculation.
My stance: any AR-pessimism from such speculation is overblown and far from indicative of Apple’s plans. There’s plenty of evidence that AR/spatial remains central to Apple’s long term strategy.
3/ Next, I (conveniently) decided predictions between just Apple & Meta would be a disservice to the industry. Niantic is certainly in the race, Snap could carve out a piece of the pie (especially amidst GenZ/GenX), Microsoft has plenty of IP and software/cloud opportunities, and the list goes on.
It’s also a misrepresentation of what spatial computing truly is (far more than just AR glasses).
See this rant to grasp what I mean.
4/ And then… the real hammer came down. OpenAI, Perplexity, and Grok all came out with their ‘deep research’ GenAI products.
These tools spit out a comparable research/analysis report within minutes. The level of depth and the predictions blew me away. They were spot on, and like most writers in the world today, I fell into a state of existential despair.
This then forced a deep inquiry… What is the point of this newsletter? Do I really want to compete head-to-head with AI? Or with all the people who are going to just copy/paste ‘deep research’ outputs and claim them as their own?
We’re about to experience a deluge of content like this; re-mixes of information that already exists across the internet (from S-1 filings, other newsletters, media outlets, etc).
I then came to my senses, which included a sense of relief; I was thankful for a (real) excuse to not finish this series.
I don’t want to write industry analysis or report the news. You also have a thousand options for this type of content. Perplexity or ChatGPT can spit out the latest/greatest within a blink, turning research and analysis into a commodity.
No, I started this newsletter because I wanted to move up the value chain; I wanted to stoke your imagination, collectively future cast, and create more optimism about the future.
Ultimately, I wanted to inspire and reveal how these new tools can make us more human not less; by increasing our capacity to feel, connect, and find meaning/purpose.
And so, I’m officially deciding to zig where most of the content world zags; classic blue ocean type of stuff (see Blue Ocean Strategy).
The precise aesthetic of these zigs will be emergent. But safe to say, it’ll include science fiction mixed with some education, opinion pieces, and philosophical musings.
This essay is the best corollary. 2042: A day in a metaverse life.
It bounces between a sci-fi ‘day in the life’ narrative and present-day explanations of the tech. And while the term metaverse has since become cringe, I do still believe in these core ideas, i.e. the power of a more experiential, contextual, collaborative, and intelligent internet. And the promise of digital ownership, internet-native value, and digital capital formation,
That essay was the most fun I’ve had writing since starting Medium Energy. It also received some of the most moving feedback.
And so, as David Deutsch likes to suggest, we’re going to follow the fun; one of the most uniquely human things we can do/feel.
Let the machines ‘think’ and process ‘tasks’. Let us ‘feel’ and process ‘experience’.
I explored this idea at length in this shorter essay: “Finding Solace in the Age of AI”.
In the spirit of letting AI do tasks, here’s a ChatGPT summary of the core thesis:
In "Finding Solace in the Age of AI," Evan Helda reflects on the rapid advancements in artificial intelligence and their implications for human creativity and self-awareness. Drawing inspiration from a 1974 interview with author Ray Bradbury, Helda emphasizes the importance of feeling over thinking in the creative process. Bradbury suggests that overthinking can stifle creativity, advocating for a more intuitive and emotion-driven approach to creation. Helda extends this idea to the modern context, proposing that as AI takes over more analytical and data-driven tasks, humans should focus on nurturing their unique capacity for emotion, intuition, and personal truth. By doing so, we can find our distinct place alongside intelligent machines, leveraging our innate abilities that AI cannot replicate
Couldn’t have said it better myself (ironically enough…)
This was in early 2023. AI for writing has gotten much better since. What was already a painfully daunting craft has become entirely overwhelming, for an entirely different reason. My most sinister inner voice calls it futile.
If AI can produce essays, news articles, and poetic prose at the push of a button, then what does it mean to be a writer? What is left over for the human creator?
These questions have been writhing within the pit of my despair. But I’ve since come to realize… This discomfort—this feeling of being outmatched by an infinite, tireless machine—is also an opportunity for clarity. It forces a deeper reckoning: Why does writing (or any creative effort) matter? What makes it worth doing?
I think the answer starts here: as AI erodes the scarcity of these skills, value accrues in the scarcity of meaning, intent, and originality.
In other words… AI's presence forces writers, artists, and creators to find the essence of why their work matters. It raises the bar.
Towards that end, and to help save me from myself, I’ve created some ‘writing tenets’ to guide this content going forward.
Perhaps you’ll find them useful as well.
The Medium Energy Tenets
1. Writing as an expression of lived experience
AI cannot have a personal history, unique struggles, or firsthand perspective.
Readers don’t just want facts—they want stories, insight, and vulnerability.
The question becomes: What can only I say?
Example: An AI could write about "the experience of loss," but it cannot recreate the specificity of your grief—the little details, the precise way it changed you. It cannot tell a story about how it felt in your body, how it changed your relationships, how you made sense of it.
2. Writing as provocation, not just information
AI excels at answering questions but struggles to ask the right ones.
The best writing doesn’t just provide information—it challenges, provokes, and invites participation.
AI may summarize Nietzsche, but it will not wrestle with a philosophical contradiction in real-time, questioning itself, evolving its stance mid-sentence.
Example: Imagine an AI writing about the impact of AI itself—it can aggregate knowledge, but can it wrestle with the existential anxiety of being replaced? Can it capture the precise feeling of watching technology erode the thing you thought made you unique?
3. Writing as a transmission of human connection/empathy
People don’t just read for knowledge; they read for communion with another mind.
The most compelling writers make you feel like you're seeing the world through their eyes.
A great book or essay is not just a collection of words—it’s a portal into someone else’s interior world.
Example: When you read Ayn Rand, you don’t just absorb facts about 1950’s-60s America—you feel her presence, her worldview, her frustration, and her hope for a better future. AI can imitate her tone, but not her being.
These tenets aren’t just for writing—they’re for anyone creating in a world where AI is rapidly automating skills and commoditizing cognitive inputs.
In music: If AI can generate endless symphonies, what makes music valuable? Live performance, artistic intent, and cultural movement.
In entrepreneurship: If AI can automate a business, what makes a company thrive? Brand, human relationships, authentic mission.
In art: If AI can paint in any style, what makes an artist matter? Perspective, rebellion, emotional urgency.
The common thread: What AI commoditizes, humans must transcend.
In conclusion: this shift is not a death sentence for creativity. It is a purification. AI eliminates the predictable, the formulaic, the interchangeable. What remains is what was always worth doing: the work that only humans, with our contradictions, desires, and histories, can produce. AI does not make writing or music or business meaningless—it raises the bar for what is meaningful.
With all that said, I feel obligated to give at least some ‘finality’ to this non-finale on the AR + AI Race.
So, here’s an easy way to consume the key ideas & opinions from my jumble of rough drafts.
In the spirit of feeling & fun, these are the takes that elicit the most emotional responses in my daily convos. These will also be fun to look back on in 5-7 years and see how things played out.
If any of these strike a chord with you too… good! Let's hear your objections or thoughts in the comments below.
Thanks for taking the time, and see you on the other side.
The AR + AI Race | Summary of Predictions
Bottom Line Up Front: If I had to pick a winner right now, from a scale and adoption perspective… my money is on Meta.
Why Meta Has the Edge
Software: AI Agents Will Redefine the OS
Apple thrived in the mobile era with direct human interaction (app UI/UX, touchscreens). But that’s not the world we’re heading towards.
We’re heading towards a world of AI agents. How much will that old iOS UI/UX matter? Do I really need to swipe & search for apps, answers, and data?
Or… can I just summon it, and get bits and pieces of visual cues if/when needed? Better yet, what if the UI changed in real-time based upon my use case and context (where I am, who I’m with, what I’m doing)? Aka a ‘generative UI’; one that could turn the world into my desktop background and display information all around.
Point being… AI agents have leveled the OS playing field. The OS and mobile app paradigm needs to be redesigned from first principles.
Meta has no legacy OS baggage and can build AI-native UX from scratch. Traditional software moats (UX/UI) are eroding—data and AI models now define competitive advantage.
Meta’s social data dominance (Instagram, WhatsApp) provides AI with unparalleled context for AR applications. Apple’s privacy-first approach may appeal to high-end users, but Meta’s ability to personalize experiences at scale will likely win the mass market.
Innovator’s Dilemma: Apple is Handcuffed by iPhone Success
Apple’s entire business is built on iPhone + App Store revenues.
Success in AR, or with ‘generative UI’ could cannibalize its core products, making them hesitant to disrupt their own cash cow (e.g. trying to use the iPhone to power glasses creates subpar tradeoffs & risks)
Meta, with no mobile OS baggage, is free to rethink spatial computing from first principles.
AI: Meta’s Unmatched Scale and Infrastructure
Meta is building the largest AI infrastructure of any consumer tech company:
600K GPUs by the end of 2024, powering Llama 3 and other cutting-edge models.
Meta’s GenAI tools (video editing, 3D reconstruction, Hyperscape) push AI beyond text/image generation into AR/VR environments.
AI-first products like Meta Ray-Bans provide real-world training data for AI assistants.
Apple, by contrast, focuses on privacy-preserving AI via on-device models—good for security but limits scalability and AI’s real-world utility.
Hardware: Meta’s Multi-Tiered Approach vs. Apple’s High-End Strategy
Meta is playing a long game, starting with the low-cost Meta Ray-Bans, then bridging to full AR glasses.
Apple’s Vision Pro is premium ($3,500) and closed—limiting adoption.
Meta’s hardware strategy is modular, with Quest, Orion (AR glasses), and Ray-Ban Smart Glasses feeding into a broader ecosystem.
Apple’s integration strength (hardware + software) is legendary, but their hesitancy to release until "perfect" may slow their AR adoption curve.
Founder Vision & Leadership:
Mark Zuckerberg is leading from the front, making high-risk, long-term bets on AI and AR. He’s just feeling himself, in full on Super Saiyan mode (if you know, you know).
This is in contrast to Apple’s Tim Cook. While Tim is playing it safe and optimizing existing product categories, Zuck is sprinting towards an entirely new computing paradigm—driven by personal conviction rather than short-term financial pressure.
Philosophy: Open vs. Closed Systems
Meta embraces openness, iterating in public with regular R&D updates, while Apple’s secrecy limits developer inspiration & engagement.
AR is too nascent to be locked down—developer feedback and rapid iteration will dictate success.
On the more technical front, open platforms (PC, Android) have historically enabled faster ecosystem growth vs. closed systems (Mac, iOS).
Developer Trust & Traction
Meta’s open philosophy and more horizontal platform approach will earn more developer trust & enthusiasm.
Meta already has a head start here. They’ve been cultivating an AR/VR developer ecosystem for years with Quest devices, while Apple’s restrictive ecosystem is frustrating devs, not to mention a lack of community/dev ecosystem investment.
Quest’s accessibility (lower price, open tools) means developers can experiment freely—Apple’s Vision Pro is a walled-garden premium product with limited sensor access.
A thriving developer community = more innovative applications, which is critical for AR’s to find product/market fit in the early days.
Fashion & Culture: Ray-Ban > Apple in the Glasses Market
Apple is a world-class design company, but Luxottica (Ray-Ban’s parent) owns the eyewear industry—from brand cachet to global retail distribution (18K+ stores vs. Apple’s 520).
Meta’s partnership with Ray-Ban positions them as the "cool" AR player from Day 1, while Apple struggles with making AR glasses fashionable.
Product Iteration & Consumer Comfort
Meta is learning in real-time with Ray-Ban Smart Glasses, refining AI + AR interactions in a natural form factor.
Early exposure builds consumer trust—Apple, by contrast, risks waiting too long and launching a product that feels unfamiliar.
Privacy concerns won’t be a major differentiator—most consumers prefer better experiences over strict data protection.
Distribution
Meta’s Luxocitta/RayBan partnership is underrated from a distribution perspective.
They have over 18,000 stores globally and one of the most recognizable brands in the game.
Apple has 520 stores. This does attract the higher/more premium end of the demographic spectrum, but I see a world where Meta, with a cheaper and more accessible product, introduces more people to AR glasses, and sooner.
Product Portfolio: Meta’s AI & Social Ecosystem Will Drive AR Use Cases
Meta controls the most engaging social networks (Instagram, Facebook, WhatsApp)—perfect for AR-driven communication and content.
AI + AR will thrive on personalized, real-time contextual awareness, and Meta’s unmatched social data enables next-gen AI-powered AR assistants.
Apple has strong AI privacy protections, but lacks the scale of Meta’s training data and ‘social distribution’ channels.
Final Verdict: Meta is better positioned to win (for now…) Key reasons being:
Founder vision (Zuck is all-in, while Cook optimizes existing products).
Philosophy (open & iterative beats closed & secretive in early markets).
Developer trust (Meta is enabling experimentation, Apple is limiting it).
AI at scale (Meta’s AI-first approach will define AR interactions).
Hardware accessibility (Meta is making AR more affordable and consumer-friendly).
Apple is not out of it by any stretch. This race is JUST beginning. There also can (and likely will) be two winners—just like iOS vs. Android. But in terms of scale and adoption, the odds are currently in Meta’s favor.