We have always worried about new forms of communication – and we shouldn't stop now

BRIEFLY NOTED

Reddit will work with third-party services to confirm the humanity of its users. Conservative activist Robby Starbuck is suing Meta after the company's chatbot claimed he participated in the Jan. 6 riots. ChatGPT's hallucinations are getting creative. Pinterest will label AI-generated content based on image metadata and its own classifiers (expect mistakes). A Sydney radio station used an AI host for six months without disclosing it to listeners. A service provider pulled the plug on MrDeepFakes, the largest distributor of deepfake porn. A joint investigation revealed that one of the admins of MrDeepFakes was a Canadian pharmacist named David Do. Grok is now an AI undresser. A popular Telegram bot let users generate a pornographic AI video with just one photo of the target. Trump Media's Truth+ is broadcasting documentaries about lizard people. Elon Musk amplified a Russian-aligned influence operation that otherwise was getting little traction. Disinformation about the Pahalgam attack in Jammu and Kashmir is proliferating. Germany's new coalition agreement includes an unclear promise to fight information manipulation.

We have always worried about new forms of communication – and we shouldn't stop now

In the spring, wildflowers carpet the hills of Southern California with colors so vivid they can sometimes be spotted from space. Much like the cherry trees blossoming on Roosevelt Island, where I work, a beautiful but routine seasonal event leads some people to lose their minds. In 2019, tens of thousands of visitors clogged the local roads of Lake Elsinore to see – and be seen in photos surrounded by – the "superbloom." They trampled poppies, uprooted flowers, and left locals bemoaning a "flowergeddon."

In his newish book Superbloom, Nicholas Carr uses this social media-induced frenzy as a metaphor for our current information ecosystem. "We live today in a perpetual superbloom – not of flowers, but of messages."

The metaphor didn't quite click for me, but that's no reason not to pick up this book. Carr situates our current information crisis in 150 years' worth of communications theory and technological progress. He is no optimist, arguing that increasingly efficient communication has "widen[ed] the gap between the pseudo-environments in which people think and the real environments in which they act."

Radio did not make war impossible

The progress of communication technology has always been accompanied by bursts of optimism and pessimism (many instances of the latter are wonderfully cataloged in the Pessimists Archive).

Charles Cooley, a 19th century American sociologist, worried that universal access to books had resulted in greater individualism as readers broke from their immediate surroundings and created ties with imagined or faraway subjects.

The 1865 International Telegraph Conference was heralded as a "veritable Peace Congress," and an 1899 New York Times editorial argued for cheap telegrams because "nothing so fosters and promotes a mutual understanding and a community of sentiment and interests as cheap, speedy, and convenient communication." The mood changed with the outbreak of World War I, when the telegraph was accused of overwhelming the "communicative art" of diplomacy with "too much communication."

Similar ebbs and flows followed the advent of radio and television.

Guglielmo Marconi predicted wireless radio "will make war impossible, because it will make war ridiculous." That was in 1912. Calls to avoid rigid legislation on radio that "would hamper the development of a great modern enterprise" sound just like today's editorials arguing for a laissez-faire approach to generative AI. This shifted after the Titanic sank and rescue missions were put in jeopardy by amateur radio users spreading misinformation. Totalitarianism regimes excelled at weaponizing mass media, tainting the radio and television by association.

We've been Zucked

Then came digital media.

In 2006, Yochai Benkler captured the zeitgeist in his argument that "the Internet democratizes" by toppling mass media and introducing a networked public sphere. In 2012, Mark Zuckerberg claimed in Facebook's IPO filing to be inspired by the printing press and television, which "by simply making communication more efficient [...] led to a complete transformation of many important parts of society."

The web and social media certainly transformed parts of society, if not always for the best. Carr concludes that Benkler's representation of the web as a place for cooperative filtering and participative civic information was "thoughtful, logical, and wrong."

At a minimum, more information didn't lead to a more informed public. The explosion in the amount of information available and the speed with which it circulates overwhelms our capacity to apply System 2 thinking. Here's Carr again:

We might have gotten a democratized marketplace of ideas. But those ideas were Pizzagate and QAnon.

Information overload and myth-making

AI chatbots complete the takeover of the information ecosystem by electronic media. First, machines replaced humans in the mere transportation of information, through the telegraph, radio, television, and email. Then, social media and search algorithms took on editorial ranking and filtering tasks. Today, content production itself has ceased to be a necessarily human activity.

The consequences of this final stage will be the subject of books yet to be written. But it isn't hard to look at the past, as Carr has done, to make an educated guess about what's in store for us.

Key figures in the early development of electronic telecommunications also believed in telepathy; the telegraph era coincided with the emergence of Spiritualism, a religious movement that believes in communicating with dead people. It's hard not to see echoes of that time in the anthropomorphization of generative AI and the obsession with chatbots resurrecting dead people.

Even if we don't end up worshipping at the altar of ChatGPT – and I'm not ready to discount that – the information overload will get worse. Virtual influencers posting AI slop 24/7 will further degrade our capacity to understand the world.

Canadian philosopher Marshall McLuhan said that "when man is overwhelmed by information, he resorts to myth. Myth is inclusive, time-saving, and fast." Carr adds in Superbloom that "a myth provides a readymade context for quickly interpreting new information as it flows chaotically around us."

Combined with the fact that "opinions emerge from affiliation, not vice versa," this opens the field for populist leaders to become "totem[s] for group identity" and shortcuts through which to interpret the barrage of information.

If there is a solution, it is to slow things down. Carr advocates – like other thinkers I respect – for "frictional design" that throws "virtual sand into the virtual works." And yet this is unlikely to happen, as he readily admits:

Reversing technological progress is moot, but it doesn't mean we can't try to steer it in a better direction.

Digital literacy expert Mike Caulfield, whom no one can accuse of naive techno-utopianism, has been working on specialized prompts that turn Claude and ChatGPT into a personal fact-checker. What I find most valuable in the output of these prompts is less the specific answers than a structured process to suss out available information.

In an ideal world, we would use the same tools that are accelerating information production to an unbearable pace to slow down how we process all that material. Instead of simplistic summaries skimpy on sources but heavy on hallucinations, we could get LLM retrieval that provides rigorous contextualization.

That is not what the tech giants are offering, but it should be what we demand.