Patterns and Predictions for 2024
Thoughts on what will have a lasting impact on us at the intersection of creativity and technology
“Trying to predict the future is a discouraging, hazardous occupation.”
— Arthur C. Clarke, 1964
What I said in 2022
In the first week of 2022, I wrote down my five thoughts for the year at the intersection of creativity and technology. I called them “thoughts,” not “predictions,” because they were observations of new patterns that would have an impact on how we create and society at large over multiple years to come. A prediction tends to be a finite event in time whereas patterns need time to emerge. Changes are gradual until they aren’t.
Let’s see how my thoughts fared since then. Or feel free to scroll down to “Patterns and Predictions for 2024.”
What I said #1: 2022 = 1999
“2022 will be like 1999. Mostly up and then a sudden drop (like the dotcom bubble back in the day), followed by 15–20 years of drastic and upward changes like we’ve gone through in the last two decades but in new and unexpected ways.”
Verdict: 🤝
At the time of the writing, NFT and the metaverse were still hot topics. The launch of ChatGPT was 11 months away. A few years earlier in 2018, OpenAI introduced the first GPT so there was an early glimpse of generative AI.
Did I think that something like ChatGPT would launch in 2022 and take the world by storm? No. Did I sense some big change was about to happen and we’d see the next wave of massive transformation like we did at the turn of the millennium? Yes. I was partially correct so I’d put my verdict as 🤝
What I said #2: New Creativity = (Making x Tenacity) x Anyone
“Creativity is now decentralized. Creative work used to belong to the elite few: those who had access one way or another (i.e. money, resources, titles called “creatives,” etc). Now, it belongs to anyone with not only the will to make something but also the tenacity to continue making.”
Verdict: 🤝
I admit that this was a broad statement to make, hence 🤝 as the verdict. The sudden rise of AI from 2022 to 2023 accelerated this drastically, way beyond what we thought was possible. We, the creative elites, didn’t see this coming.
At the heart of digital—devices, software, and the Internet— is the power to take what was for the elite and make it accessible to everyone. The power of AI is that not only does it eliminate the barrier to access but also to skill. You now don’t need much skill to create.
What I said #3: Creators > Platforms
“The power is now shifting from Platforms (i.e. Facebook, Instagram, TikTok, etc) to Creators (not “creatives” or “creative elites,” see above). Platforms are becoming homogenized and formats are increasingly the same (i.e. Snaps, Stories, Shorts, etc), making platforms less unique. By 2025 (if not already), many Creators will be their own Platforms.”
Verdict: 👎
Platforms still hold power over Creators and will continue to do so. One would have to be a mega-celebrity to be Platform independent so this statement wouldn’t apply to 99% of individual Creators. Everyone has and needs their primary Platform of choice.
What I said #4: NFT = Opportunity x 0.05
“95%+ of NFT-related businesses today will be gone by 2030 or sooner, just like the dotcom companies of the late 90s that flamed… There will be a lot of [NFT] tools and businesses with no value.”
Verdict: 👍
My prediction here was generally right but too conservative in my time estimate. The term NFT reached its peak popularity on Google in 2022 and has now faded into our rearview mirror. It didn’t need several years. It just took a few months for people to move on.
What I said #5: Big Data + Profit = RIP
“If the last decade has taught us anything, Big Data in the name of profit was the wrong pill to take. Data needs to be used for the betterment of humanity, not growth alone. Some companies would argue that they are using data to “connect humanity,” blah, blah, blah. Exactly.”
Verdict: 👎
While the laws around data privacy are getting stricter, with the EU leading the charge, generative AI relies on a massive amount of data to work. Companies may not be able to use third-party data as freely as they used to but they will aim to shift towards first-party data for their businesses. Data continues to be the oil of our times.
Patterns and Predictions for 2024
1. Do more better with less
First, about doing more with less.
This isn’t a new pattern. This is what technology does to humanity and society. Technology always raises the expectation that we could or should do more with less: Less money, less time, and fewer people.
The introduction of technology that automates something allows us to do more with less. We can go all the way back to the First Industrial Revolution in the 18th century to witness this tendency.
One could argue that we are entering the Fifth Industrial Revolution with the advancement of generative AI.
What makes this Revolution different from the previous four Industrial Revolutions is twofold.
One is the speed at which progress—or a decline—is happening. If we look at history, the intervals between the Revolutions are becoming shorter each time. The difference between where AI was 18 months ago and today is staggering.
Two, before new types of work can be industrialized, AI will deindustrialize various types of work, like creative tasks previously reserved for humans. As this Financial Times article indicates, “the gig-worker model of performing one task for multiple clients — copywriting or logo design, for example — is especially exposed.”
So it’s obvious to think that we need to do more with less. This will be truer as we move deeper into the New World Order of AI.
One study this year by researchers at Washington University in St. Louis and New York University, cited in the article above, indicates a noticeable and immediate impact on the freelance market. Before the introduction of ChatGPT, the percentage change in the monthly freelance jobs on Upwork, a large online marketplace for freelance jobs, was between -0.5% and +0.5%. In other words, it was more or less flat.
After the launch, the study shows, the percentage change dropped to almost -3% within five months. When it comes to earnings, freelancers on Upwork were reporting that their earnings dropped by about 10% in the same five months. That is a significant drop in such a short amount of time.
The challenge with the “do more with less” mentality is that it could be a race to the bottom.
In an age where we will be seeing an avalanche of synthetic and mediocre content, quality will be the differentiator. We need to race to the top.
Humans will need to pivot to quality from quantity. In other words, do better, not necessarily more, with less.
Prediction: One in ten creative and marketing professionals will need to pivot. Or, they will need to have the #OpenToWork badge on their LinkedIn profile.
According to data from Interviewing.io, an interview practice platform and recruiting marketplace for engineers, only 1.4% had that #OpenToWork tag in 2021 compared to 4.2% in the first quarter of 2023. I suspect this is more if you focus on the creative and marketing industry.
The number will continue to rise to about 10% for the next 6 to 12 months before the trend reverses.
2. The 10-80-10 Rule of AI
This sequence of numbers refers to “10% human at the beginning, 80% machine in the middle, then 10% human to finish.”
It’s not hard to imagine that one of the first areas in which we will see a major change forced by generative AI is content creation, as we are starting to see already. Within this large sector, I believe there are two sub-industries to which this 10-80-10 rule can apply immediately and profoundly: performance marketing and animation.
Hikari Senju, the founder of Omneky, told me about this rule last month in San Francisco. It’s a Silicon Valley startup that recently raised over $20 million to build an AI platform that generates, analyzes, and optimizes personalized ad units at scale. He founded the company in 2018 so it’s been around even before generative AI became the buzzword it is today. Think of this platform as a banner ad generator that can spit out tens and hundreds of variations within minutes and test them. Forget A-B testing. Omneky can do A-to-Z testing and optimize those ad units in real-time.
The good news is that Hikari aims for his platform to do 80% of the work humans used to do—and didn’t necessarily enjoy—but with more quantity, efficiency, and scale. AI is well suited for this kind of performance-driven advertising.
The bad news? Notice I didn’t say “quality” in the last sentences.
Another company that I’ve come across this year is iKHOR Labs in Los Angeles. The company builds AI-powered professional animation tools that help directors and animators do in-betweening, keyframing, line art finishing, and auto-rotoscoping.
Prediction: By 2025, 75% or more of performance advertising units will be generated by AI.
iKHOR will be bought by Disney or Netflix by the end of 2024.
Sometime in 2025, one of these companies will release a feature-length animation film that follows the 10-80-10 Rule of AI.
3. Interface shift
Sam Altman, the CEO of OpenAI, has said that he was somewhat embarrassed about the quality of ChatGPT when it was being released. Quite the contrary, we all thought it was surprisingly good.
As such, this is causing the interface to shift in various categories such as search, commerce, and automotive.
Search
First and foremost, this interface is changing how people find and access information. For the past 25 years, the search interface has been the primary window. As all of us have experienced, getting answers from ChatGPT, Bard, or Perplexity AI, is much easier than searching, precisely because we no longer have to search. Answers can now be presented to us. The hallucination problem might not go away completely but that answers from these AI tools will be acceptable 95 to 98% of the time. That’s good enough.
The point here is that the very notion of “search,” which governed the flow of information, is now getting challenged fundamentally. The way the traffic is generated on the Internet will change in the next 12 to 24 months.
Good news for the users, and bad news for publishers who survived based on links, clicks, and traffic.
Commerce
The term “conversational commerce” started to proliferate on blogs of Shopify and Zendesk as well as thought-leadership articles by Accenture and Forbes.
All of these opinions claim that conversational interfaces like chatbots, messaging apps, and voice assistants are the future.
I happen to think otherwise: Conversational commerce is bullshit.
Just imagine when you want to or need to talk to a store associate in a store. In the context of shopping, conversations are useful when you encounter a problem, i.e., “I can’t find ___.” A good shopping experience, in most cases, eliminates the need for that conversation.
Shoppers buy more in physical stores than online because they wander and discover, not search. The next interface shift powered by AI in commerce won’t be more chat bubbles. Changes will be more subtle on the surface but what shifts is behind the interface and how you discover.
Commerce companies that leverage AI to allow for more contextual discovery will win.
Automotive
Another example of the interface shift will be in the automotive industry.
There are autonomous taxis on the streets of San Francisco today. However, the day when drivers can take their eyes and hands off their wheels is still a while away.
Take elevators, for example. By 1900, completely automated elevators were available, but passengers were reluctant to use them. It would take almost 50 years for automated elevators to become mainstream in the US. This report suggests that mass adoption of autonomous vehicles remains at least a decade away. It could even be two.
That then gives a big window of time for the interface within cars to evolve for the safety of drivers. The potential of ChatGPTs and alike is to allow our eyes and hands to be on the road and the wheel and make our voice and ears the primary input for the next generation of intelligent interfaces for drivers.
Prediction: As much as I want products like Humane’s AI Pin to succeed, it’s painful to watch its founders, on a recent news show, try to explain and struggle to answer simple questions like what it does or why it’s better than what we already have. By Christmas 2024, its price will be cut in half.
On the other hand, Apple’s Vision Pro will succeed, not because millions of people will flock to buy it but because it’s $4,000 apiece. Apple needs to sell only 250,000 units to reach $1 billion. Not a high bar for the world’s most valuable company. Some will buy for its novelty factor but it won’t be enough to make AR/VR the next interface.
While Apple’s been rumored to be working on a vehicle since 2014, we won’t hear about an Apple car in 2024. Why? See above. They don’t want to crowd the airwaves with another major product.
To summarize, here are the Patterns and Predictions for 2024:
Do better with less
1 in 10 creative/marketing people need to pivot in 2024. Otherwise, #OpenToWork.
The 10-80-10 Rule of AI
AI dominates in performance marketing.
A feature-length animation film is made with the 10-80-10 Rule of AI by 2025.
Interface shift
A major shift in search, a move to contextual discovery in commerce, safer opportunities in automotive
AI Pin fails. Vision Pro doesn’t. No Apple car just yet.
Thanks for reading. Please hit reply if you have comments, questions, and feedback. I’d love to hear from you.
This will be the last essay for 2023. Thanks for being a subscriber.
Looking forward to seeing you here again in 2024.
Contact lens for the AR world.
These predictions (and your honest and revealing look at last year's) are great! I might add one more under your Interface Shift insight: I think GPT and LLMs themselves are in a "command line" phase, and that we'll start to see other interfaces to interact with generative AI tools. At least that's my hope.... What do you think? Is it coming, and would it be an improvement?