See it Coming
We can't predict the future. We can better anticipate and prepare for it.
"There is nothing in which people differ more than in their powers of observation."
- John Burroughs
In 1965, Alvin Toffler coined the term "Future Shock." It's a state of shattering stress and disorientation from too much change in too short a time.
Toffler likened the future-shocked person to a soldier or disaster victim. They would face extreme change and unanticipated events — a “fog of war” in everyday life. This state of mind helps describe the sense of chaos and erratic behavior we see today.
The accuracy of Toffler’s predictions from the mid-1960s is notable. The odds of successfully predicting human or market behavior are long. Complexity theorists, strategists, and future thinkers constantly miss unexpected cultural breaks. Big things like Brexit, Trumpism, COVID repercussions—even Elon Musk's purchase of Twitter.
The degree to which “unthinkables” shock us is worth thinking about. Are we as adaptable to change as experts believe? Is it possible to see social undercurrents before they tip? Is it a fool's errand to predict future events?
We can’t predict precisely when societal disruptions occur, but we can plan and model plausible scenarios. A forward-leaning perspective is critical if we’re to benefit from astounding technological advances and not be psychologically destroyed by them.
Below is a framework I use to help see societal disruptions form — in particular the jarring, permanent dislocations resulting from technological change. The progression from start-up invention to societal break has a logic behind it.
This model is based on social tech consulting going back to the early 2000s and ongoing media research. It also leans on diffusion models from those much smarter than me — Everett Rogers, Geoffrey Moore, Grant McCracken, and Carlota Perez. Another great teacher, Clay Shirky, said things don't get interesting until the technology (i.e., novelty) gets boring. Meaning when full "deployment" takes hold, societal dislocations happen.
From my interpretation, technologies that pass four "gates" lead to paradigm shifts:
1: Inception. New ideas supported by early-stage investment get into the public bloodstream. Founders hypothesize about solutions but pivot continuously to find “product-market fit” in the real world.
2: Use case expansion. The ways people will use new technologies are unknown to the entrepreneurs building them. Craigslist started as an email among friends. Instagram began as a whiskey app. Slack started as a video game. All defied convention and eventually created a new language, categories, and use cases.
3: Defining financial events. Over time, the value gets locked in. In the most prominent cases, growth in users and clarity of value lead to significant financial events. These include late-stage funding, acquisition, or public offerings. When the economics change, offerings do too. Social media networks becoming predominant ad networks is a case in point.
4: Full deployment. Technology that reaches mass adoption changes society. In doing so, it defies previous cultural logic. Those who watched Twitter's debut at SXSW in 2007 could not have predicted that it would aid societal revolutions. That it would have an impact on elections. That it would serve as the president's bully pulpit. That it would lead to open-source military intelligence. It has become a dividing line in the debate over free expression.
Twitter follows this societal disruption model end to end. The full inculturation of Twitter as an information gateway and free speech fulcrum is why Musk just paid $44 billion for it.
Let's look at another example in progress.
Earlier this month, The New York Times published a Steven Johnson essay asking, "A.I. is mastering the language. Should we trust it?"
The essay is a decade late. Automated text generation is a field of incredibly fast and consequential development.
Let's look at its history with this cultural framework in mind:
1: Inception. In 2013, 17-year-old Nick D'Aloisio sold his budding tech company called Summly for $30 million. Its algorithms generated automated summaries culled from thousands of news sources. Yahoo! acquired it to help source and summarize news content on its array of platforms. Summly was one of a number of start-ups focusing on data mining, autosuggestion, and summarization without the involvement of humans.
2: Use case expansion. Other firms, such as Narrative Science, Automated Insights, and IBM/Watson, followed suit, resulting in the formation of a new category. Natural language processing (NLP) evolved into Natural Language Generation (NLG). Elements of personalization, customer service, and media production would all be automated by developers. Others, such as Microsoft, built out 'Turing Labs' to make machines "more literate" in both thinking and output.
3: Defining financial events. In 2015, high-profile tech execs, including Musk and Sam Altman announced OpenAI. They pledged over $1 billion to build "beneficial" artificial intelligence. OpenAI would announce a series of text generation models (GTP, GTP-2, GTP3) trained on trillions of words from the Internet. To put it in context, GTP-3 and similar models from Google, Meta, and DeepMind train on the entire Internet. Early trials included imaginary conversations between historical figures, movie summaries, and writing code. The developer community's response was equal parts awe and surprise.
4: Full deployment. While not fully deployed, NLG development is happening at an astounding rate. Over 300 different GTP-3-powered applications became available in the first nine months of API access. Other models, such as China’s WuDao 2.0, train on more extensive data sets than GTP-3. General-purpose apps can now write legal documents, original ideas, articles, blog posts, and digital ads.
This is the tip of the iceberg. Think of data that can tell its own stories. Software that writes itself. Auto-generating adaptations to news and event coverage. Now OpenAI’s DALL-E2 can create images and art from simple descriptions.
It's impossible to say where this will lead us. The ramifications of NLG are comparable to those of super-hyped categories like Web3 and the Metaverse. It doesn't even come close to the buzz. In fact, according to Google Trends, GTP-3 isn't even on the map.
So what happens when NLG technologies progress into full "deployment?"
A significant elevation in intelligence generation and automation is already underway. Systems like GPT-3 could soon replace Google or Wikipedia as our primary resource for new knowledge.
On the other hand, the list of considerable downsides is long. Job loss for those paid to research, write, and code. Out of control bias. Nuclear-grade weaponization of disinformation. Bots overwhelming social media discourse. Inability to distinguish between humans and non-humans. Online identity in constant question. The end of the Internet as we know it.
GTP-3 must be on our societal radar. We aren't far from the day when GPT-3's commercialized offspring begins to swarm our digital discourse. How people react is where real disruption takes place.
We can’t predict exactly how or when it will happen. We can see directions and try to blunt use cases that pose an existential risk. That requires a collective will, models to track developments, and where necessary, regulation.
With disruptions like GTP-3, we can choose to anticipate and take full advantage of the potential. Or choose not and find ourselves behind an 8-ball of events where we no longer have control.
Marshall McLuhan said that technology is put out before thought out. That latter requires effort to see and anticipate effects before it's too late.
Given what’s a stake, this added energy perspective is well-worth thinking about.
*I used Grammarly, Hemingway, and A.I. Copywriter to write this. I cannot imagine writing without this automated assist.
FINAL NOTE
Who could have guessed that platforms would become larger, and some would argue, more powerful than many countries? And even more improbable, that it took less than 20 years to happen. Artificial intelligence investment reached $77.5 billion in 2021 alone. So, how might disruption manifest itself again in the next 20 years?
We can only guess — or monitor it with frameworks like the one used in the post.
Click through here to see the immensity of platforms on the global economy and society. As AI takes hold, we’ll likely see an even greater impact and accumulated power in the future.
Perspective Agents are things that push the boundaries of thought.
If you found this post to be useful, I would be grateful if you shared it with your contacts.
And if there are topics of interest to you, drop me a line: cperry248@gmail.com