The technology of the late 19th and early 20th centuries is filled with neglected but interesting bits. One that I come back to frequently is the steam-driven airplane. While the idea seems an obvious non-starter today, this was an active line of research which produced some surprisingly successful aircraft.
The most impressive of these were the unmanned aircraft of Samuel Pierpont Langley, which made multiple flights three to five times further and eight years earlier than the Wright brothers. Langley's plane may now be best known due to a slanderous TED Talk from a hack motivational speaker, but it was a marvelous piece of engineering.
Here's how Scientific American put it in March of 1904.
In 1896, for the first time in history, a mechanical structure, free of any attachment to the ground and wholly without any supporting power but its own engines. made several flights of over one-half mile each. Mr. Langley had at this point reached the original aim of his researches in this direction---that of demonstrating, as a question of mechanical engineering, first. the conditions for, and second, the possibility of accomplishing, mechanical flight.
He was remarkably close to building the first working manned aircraft, and it's fun to speculate, in an alternative-history kind of way, about how things might have been had events broken more his way. But in terms of the history of aeronautics, it would have made little difference. Internal combustion was the future of flight. Steam was a dead end.
The Wright brothers' plane was the very opposite of a dead-end technology. The basic principles and design choices were all completely sound, and you can trace a fairly direct line from those first models to the passenger planes and military aircraft of two or three decades later.
That said, for all the excitement, no serious person looked at this and said this is commercially viable technology. As with Edison’s phonograph, which had also shocked the world 30 years earlier, while virtually everyone recognized this as a breakthrough, it was also clear that the technology would have to evolve considerably before it could be rolled out for widespread business or military applications.
On his seminal album The Button-Down Mind, Bob Newhart imagined a conversation between the Wright brothers and a post-war era corporation trying to monetize their breakthrough. The humor of the monologue came partly from the absurdity of trying to stack multiple passengers on the wing of the Wright Flyer or making a coast-to-coast trip taking off and landing every 105 ft, but much of it also came from the banality and shortsightedness of 60s-era corporate culture in the face of a stunning, world-altering step forward. It’s a comparison that’s, if anything, even sharper in the age of venture capitalism.
Which brings us around to the original question. Are LLMs a steam airplane—a wonderful piece of engineering and a major advance, but still a dead-end technology doomed to be pushed aside by something better before it makes its mark?
Are they Newhart’s airline—a viable and important technology that isn’t ready yet to support the commercial applications that people are trying to impose on it?
The mountains of money that are being poured into AI in 2025 are mind-boggling, and if either of these possibilities turns out to be true the economic implications are stunning.
I’m inclined to believe that one of these two possibilities is true (leaning more toward Langley than Newheart) , which would be very bad news for a lot of people—perhaps, depending on how it plays out, for most of us. Obviously that’s just an opinion, but given the stakes, these questions would seem to be worth asking.
As you might guess, my opinion is that llms are neither of those: they're an inane parlor trick.
ReplyDeletellms don't have a world model to check their results against. They can't "reason" because they have no facts to reason with. They output randomly generated text and force the user to do the reality check. If the output's bad, you are supposed to rewrite your prompt until you find a prompt that elicits desired output. (This is very much how psychics work.)
Sound recording's an interesting technology case in that it was so desperately needed* that it became a monster hit well before it was ready for prime time. By the early 1950s, it was good enough to provide lovely listenable music, but before that it was painful.
(The bebop and hard bop of the 50s and 60s was developed jn the 1940s, but those recordings are horrible. Sigh.)
*: My father survived WWII by wearing out Benny Goodman's recordings of the Mozart clarinette concertos.