Who is the authoritative voice of reason for the effectiveness of any educational approach or tool? It takes years to sort things out and by then the approach of the day has moved on. Education is an area where everybody feels like they are an expert. In the US, we can't agree on the simplest things impacting our children.
Clearly it should, at least, never be the company that stands to profit from selling said technology to the government. Like you I am highly skeptical, the idea of classrooms is an idea that is over 5,000 years old. It's kinda been nearly perfected at this point where pedagogy mostly points toward self-sufficiency as soon as possible.
Also think it's extremely unethical to experiment on school children using tools or methods that can be detrimental to their development (see chromebook debacle, moving away from phonics). We kinda know what works very well in a public education setting, the constant experimentation does not help.
Experiment on adults that can consent, not children.
> It takes years to sort things out and by then the approach of the day has moved on.
What? Good approaches, definitionally, have quite a lot of staying power. This is true in most fields but especially in education where, it turns out, we don't really figure out net new approaches that are much better than the last ones very frequently.
New approaches are tried in education all the time. What constitutes "good" is subjective and might take years to really play out. Education is like economics... so many variables and changing starting conditions
Dolly Parton was an incredibly savvy business person, talented songwriter and performer, and an outstanding human being. One in a billion.
I will always remember her appearance on the Late Show with Steven Colbert where she moved him to tears with an a cappella song: https://youtu.be/R91VssUiGiY
I was somewhere recently where the World Cup games were on the TV, but the current score was also on what appeared to be a split-flap board mounted to the wall. Seeing your post had me look up commercial availability of these (it seems they exist, but are pricy). Serving up baseball scores in that kind of display would be pretty cool.
I always loved the story of the "three-sided" Monty Python record, where the B side had two parallel concentric grooves, causing different tracks to play depending on where the needle was dropped. I always wondered what kind of equipment went into producing it.
From what I know about record manufacturing, creating the lacquer master would involve adjusting the angle of the groove to allow for two concentric grooves to be laid, along with some care in creating that master. But once it’s done, the manufacturing process is handled through a stamping process (so a vinyl record isn’t cut, it’s pressed with a metal die that’s created from the lacquer master through electroplating).
You’d cut it normal equipment, with very wide groove spacing. Start the two grooves 180 degrees out of phase. They’ll never intersect.
You could do it with more than two grooves, just to having them at 360/n degrees apart. You’ll just have to make the groove spacing wider as the number of tracks go off. Of course that comes at the cost of playback tine.
When I was a kid, we had a "record producer" board game and the randomizer was a 45rpm record with three concentric grooves. You did a needle drop and it said It's a Hit, It's a Flop, or Break Even.
De La Soul's 12" Single for "Me Myself And I" also has its second side cut with two grooves. The hype sticker says, "3 Sides." Each time you put the needle down you have a 50/50 chance to hear the song you're trying to hear. :-)
The same goes for the 1994 first pressing of Marillion's Brave.
Side 4 has a double groove, which would give you either The Great Escape + Made again (a sort of a happy ending) or The Great Escape + 20 minutes of water sounds (which can be interpreted as the sad ending).
> Yes, the AI may have produced the recommendation but a human decided to follow it, so that human must be held accountable
It is common and a mistake IMO to rely on the AI as the sole source for answers to follow-up questions. Better verification would have humans sign off on the veracity of fundamental assumptions. But where does this live? Can an AI model be trusted to rely on previous corrections? This seems impossible or possibly adversarial in a public cloud.
That has been the case for a vast swathe of time across history. It hurts because we had a nice couple of decades where it seemed that, not only was this not the case, but that we were directionally accelerating away from it.
I love the accessibility and diversity of large city living in the US, but it is definitely the exception to the rule. The US is hoping for technological breakthroughs in self driving electric cars to bail us out from the sprawl we've created.
