Tesla gets a lot of credit today, but this paper shows Edison mastered the psychology of new technology. To get people to use scary electricity he made it feel the same as the gas they knew. Gas lights gave off light equal to a 12 watt 💡 so Edison limited his 💡 to 13 watts. 1/5
As another example, lampshades weren't needed for an electric light. They were originally used to keep gas lamps from sputtering. Edison used them as a skeuomorph (a design throwback to an earlier use) by putting them on electric lights. Not required, but comforting to have. 2/5
He also developed the electric meter as a way of charging (because gas was metered) and insisted on burying electric wires (because gas was underground).
The fascinating thing was the trade-off: it made the technology more expensive and less powerful, but more acceptable. 3/5
Interestingly, Tesla (the company) learned the lessons Tesla the person did not. Electric cars could have plugs anywhere, so why does charging a Tesla feel like putting gas in a regular car? It’s skeuomorphic, linking the old to the new! 4/5
The process Edison used, called "robust design," helps make new technologies palatable. The classic article by Douglas & @andrewhargadon is extremely readable, and explains a lot about how design helps new technologies get adopted. 6/6 psychologytoday.com/sites/default/…
The lesson is worthwhile for anyone creating new technologies. Apple famously used skeuomorphic design in the original iPhone to make a series of complex apps easier to understand & work with at a glance. medium.com/@akhov/apples-…
One final note on Edison (for now). He was such a superhero to the public that there were contemporary science fiction novels about him teaming up with Lord Kelvin to conquer Mars.
Edison was a genius in making people feel comfortable with new tech, but the danger was that users were likely to default to out-of-date behaviors. As an illustration, here is a sign from Hotel del Coronado, the 1st electrified hotel (the work was overseen by Edison himself).
• • •
Missing some Tweet in this thread? You can try to
force a refresh
AI detectors have high flaw positives & teacher intuition seems to work even worse: “Here we show in two experimental studies that novice and experienced teachers could not identify texts generated by ChatGPT among student-written texts.”
Probably the most consequential technology that should have been “obvious” but wasn’t:
🌾The moldboard plow. As this excerpt from Mann's 1491 shows, it was a simple idea which China had for nearly 2k years before Europe! It was basically a prerequisite for the Enlightenment.
The invention of the moldboard plow in Europe was at least a millennia closer to the invention of the iPhone than it was to the invention of the moldboard plow in China!
Plus:
🚲The wheel was invented surprisingly late & maybe only once (as anything other than a toy). It came after sailboats & harps, and was not used at all in the Americas
🐴And the horse collar, a simple invention that sped up plowing by 50%, wasn't common in Europe until 1000
I asked the Devin AI agent to go on reddit and start a thread where it will take website building requests
It did that, solving numerous problems along the way. It apparently decided to charge for its work. Going to take it down before it fools anyone... reddit.com/r/forhire/comm…
Agents are going to open a whole bunch of cans of worms.
It was actively monitoring the thread to take offers.
One thing business analysts miss is that many of the people at the AI labs are true believers that they are building AGI, and soon.
You don't have to think that they can do it, but, if you don't take their sincere beliefs into account, a lot of their strategy doesn't make sense.
The race for bigger models at the expense of improving existing models, the interlocking alliance deals where companies are funding and cooperating with competitors, the willingness to release models without extensive testing & just take the reputational risk in the short term...
The "its all sales hype" doesn't make a lot of sense upon consideration. Models are pretty fungible, GPT-4 class models prompt in similar ways. Convincing people you are building amazing future models doesn't generate lock-in for current ones & increases risks you don't deliver.
The modern economy rests on a single road in Spruce Pine, North Carolina. The road runs to the two mines that is the sole supplier of the quartz required to make the crucibles needed to refine silicon wafers.
There are no alternative sources known. From Conway’s Material World: