Yes but man the degraded democratic public impact of this (per @himself.bsky.social) makes one shudder.

Sure, but the admins in higher ed pushing AI into all facets of university life probably have that perspective. (And I'd gently note that you seem ok assuming motives on the part of detractors that aren't really established either, such as water concerns.)

It’s just that no one was saying we needed to reorder all of society, or at least the modern university, because of Excel. That’s the irritant.

Please don’t trust your chatbot for medical advice Four separate studies all point in the same direction

"Please don't trust your chatbot for medical advice," my friend @garymarcus.bsky.social pleads, pointing to the wide range of empirical evidence indicating just how unreliable these tools are for medical diagnoses. Frustration with our medical system does not mean AI is the solution.

Great, thanks for sharing!

I highly recommend Tense Present, an essay in Harper’s by David Foster Wallace. There’s a discussion of why he teaches Standard White English that is very thought provoking.

Whose presentation is this, may I ask?

Interesting indeed, though odd to write an entire essay suggesting the historiography of science over past 60 years validates Kuhn’s description of science m and then have that paragraph at the end hand-waving away Kuhn.

Agreed on that! Less certain that people have even a basic idea of how these models do what they do.

I’m not sure who or what you’re arguing against my man. The quote isn’t about that.

My counterpoint: helping people build a basic mental model of how these tools function can make them (the people) a lot less inclined to trust them (the tools). My experience is that this is effective in moving the AI curious to become AI skeptical.

I am quoting from the linked essay.

“One of the assumptions about AI tutors is that once we replace the human with an AI agent, everyone will keep interacting with the AI as if they were still dealing with the human.

That’s a silly assumption, particularly in a school setting. Students do not even treat humans like other humans.”

“LLMs are an enterprise technology that may make certain kinds of data organization easier, or coding faster….Making it easier to write code doesn’t change that I don’t want to write code. I have other hobbies!

The actual use for LLMs in most normal people’s lives is cheating on schoolwork.”

“Frontier labs are investing astounding sums of money in mastering [coding] while starting to push into…specialized domains, such as accounting, law, healthcare, etc.

We have very limited evidence on the true balance of capabilities of these newer domains.”

So much seems to be riding on this.

I would provide a space.

Ok fair, I meant “the unknown particularities of how a specific technology will be used in the future,” I agree education generally is preparation for the experience of life.

Nice, @mpolikoff.bsky.social offers the sane yet somehow still controversial observation that AI is by and large an educational hazard, not enhancer. Not sure his suggested mitigation strategy will work but I’ll take the partial victory for reasonableness.

Recently senior executives at OpenAI and TMSC (major chip foundry) have declared the prompt engineering/generative AI phase over, and that agentic AI is where it’s at. No one knows what “the future” holds and we should stop pressuring educators to prepare students for the unknown.

Oh nice, I like him. But agreed on your broader wariness.

Wait who did DeepMind hire? They already have Murray Shanahan!

Note: Article linked above discusses the market shift by Big Tech AI purveyors in gernal. The specific quote is from TMSC's recent earnings call, and the prepared remarks can be found here via this ugly link: investor.tsmc.com/english/encr...

Anthropic, OpenAI, Google, and Microsoft agree that the harness is the product. They disagree on the price. Anthropic, OpenAI, Google and Microsoft split on AI agent harness pricing as Anthropic charges $0.08 per session hour and OpenAI ships open source.

"The shift from generative AI and the query mode to agentic AI and the command and action mode is leading to another step up in the amount of tokens being consumed." -- CEO of TSMC (chip maker)

This echoes what OpenAI's de facto COO said two weeks ago. I think it's a very big deal.

“Artificial Intelligence systems increasingly shape and permeate our mentality and social environments….relationships are reduced to functional responses. Dear friends, you, however, are real persons! Creation itself has a body, a breath, a life to be listened to and safeguarded.” - Pope Leo

To be fair that describes the President

And have shown a predilection for cutting corners already!

We might need to reboot the whole higher ed system given the way things are going.

Someone named Christie Williams has coached thousands of students how to mint a worthless college degree. “Why wouldn’t you do that?” Williams asked. “It’s kind of a no-brainer if you know about it.”

Kind of a no-brainer indeed.

Which you’d think would be one problem they wouldn’t have…

Wow I can feel the mass unsubscribing happening in real time.