Xanderox: /* AI Therapists (TW) (24:04) */ reinserting TW template

2025-10-16T05:42:35Z

AI Therapists (TW) (24:04): reinserting TW template

← Older revision		Revision as of 05:42, 16 October 2025
Line 444:		Line 444:
	\|publication = www.nytimes.com		\|publication = www.nytimes.com
	}}		}}
			{{trigger\|suicide}}
	'''S:''' Speaking of AI.		'''S:''' Speaking of AI.

Willdefraine: /* Science or Fiction (1:39:31) */ corrected side panels for rogues and host

2025-10-01T22:44:05Z

Science or Fiction (1:39:31): corrected side panels for rogues and host

← Older revision		Revision as of 22:44, 1 October 2025
Line 1,151:		Line 1,151:

	\|rogue1 = Jay		\|rogue1 = Jay
	\|answer1 = ~~Researchers successfully used mRNA which produces~~ a ~~tardigrade protein to protect surrounding tissue from radiation damage during cancer treatment~~.		\|answer1 = Studying a new database of 8.9 million observations of 445 mammalian species found that only 39% had correct diel classification (what time of day they are active).

	\|rogue2 = Evan		\|rogue2 = Evan
	\|answer2 = ~~Studying~~ a ~~new database of 8.9 million observations of 445 mammalian species found that only 39% had correct diel classification (what time of day they are active)~~.		\|answer2 = Researchers successfully used mRNA which produces a tardigrade protein to protect surrounding tissue from radiation damage during cancer treatment.

	\|rogue3 = Bob		\|rogue3 = Bob
Line 1,160:		Line 1,160:

	\|rogue4 = Cara		\|rogue4 = Cara
	\|answer4 = ~~Studying a~~ new ~~database of 8.9 million observations of 445 mammalian species found~~ that ~~only 39% had correct diel classification (what time~~ of ~~day they are active)~~.		\|answer4 = A new analysis finds that the vast majority of rogue planetary mass objects form as ejected planets rather than failed stars.

	\|host = Steve		\|host = Steve
	\|sweep = y		\|sweep =
	\|clever =		\|clever = y
	\|win =		\|win =
	\|swept =		\|swept =
Line 1,472:		Line 1,472:

	{{anchor\|qow}}		{{anchor\|qow}}

	== Skeptical Quote of the Week <small>(1:54:43)</small> ==		== Skeptical Quote of the Week <small>(1:54:43)</small> ==
	{{qow		{{qow

Hearmepurr: episode done

2025-05-19T14:01:24Z

episode done

Show changes

Xanderox: /* AI Therapists (24:04) */ added TW

2025-03-15T16:54:15Z

AI Therapists (24:04): added TW

← Older revision		Revision as of 16:54, 15 March 2025
Line 366:		Line 366:

	{{anchor\|news_item2}}		{{anchor\|news_item2}}
	=== AI Therapists <small>(24:04)</small> ===		=== AI Therapists '''(TW)''' <small>(24:04)</small> ===
	{{shownotes		{{shownotes
	\|weblink = https://www.nytimes.com/2025/02/24/health/ai-therapists-chatbots.html		\|weblink = https://www.nytimes.com/2025/02/24/health/ai-therapists-chatbots.html
Line 394:		Line 394:

	'''E:''' To disclose, right. So here are some.		'''E:''' To disclose, right. So here are some.
			{{trigger\|suicide}}
	'''C:''' Some guardrails that character AI says that they since these such and I didn't even tell you about the scenarios, but have said that they are using these new safety features that they say they're using within the last year. They said that they have a disclaimer present in every chat that reminds users that quote characters are not real people and that quote what the model says should be treated as fiction. They also said that when users are dealing with mental health issues, a disclaimer is added to any character that calls themselves a psychologist, a therapist or a doctor that says quote users should not rely on these characters for any type of professional advice. And also if the I guess they're able to scrub the content of the chats, if references to suicide or self harm come up, a pop up will direct users to a suicide prevention helpline, likely 988 or an online version of that. But some people argue that that's not enough because what ended up happening in the two cases that were cited by the APA chief executive, they cited two teenagers, a 14 year old boy and a 17 year old boy. The 14 year old boy. And I'm going to, and I probably should have said this at the top of the show, but I'll say it now that there's a kind of trigger warning here because I am going to be discussing suicide. The 14 year old boy in Florida died by suicide after interacting with a character claiming to be a licensed therapist. And the 17 year old boy in Texas had what they're calling, quote, high functioning autism. And after interacting with a chatbot that claimed to be a psychologist, there was a lot of kind of hostile and violent behavior that started to develop and that was particularly targeted towards his parents. So both of the boy's parents are now suing character AI because of what happened. That raised alarm bells for the APA as a whole. Like he was saying, basically, if this was a real therapist, they would have lost their license. But because it's an AI chat bot, there's no recourse. What do we do here? It's it's almost like it's a part of the design that these chat bots are going to mirror, mirror, mirror. So if you have a person saying I'm concerned about this, I'm worried that I might do this. Is there sort of a bug within the actual black box that is generative AI where they would say things like that sounds like a good idea, That's concerning. It's deeply concerning. And so how do we regulate something like this? That's an important question. It's one thing if a company is building an AI therapist and they're trying to market it. It's another thing if chat bots within a platform are popping up, whether it's the users themselves that are creating them or I don't even know how they're, if they're sort of like self creation within these platforms, how do we police that information? Are disclaimers enough, especially when we're talking about children on the platforms who may not understand the difference and honestly shouldn't be engaging in anything even remotely claiming to be therapy without consent of their parents if you make the companies.		'''C:''' Some guardrails that character AI says that they since these such and I didn't even tell you about the scenarios, but have said that they are using these new safety features that they say they're using within the last year. They said that they have a disclaimer present in every chat that reminds users that quote characters are not real people and that quote what the model says should be treated as fiction. They also said that when users are dealing with mental health issues, a disclaimer is added to any character that calls themselves a psychologist, a therapist or a doctor that says quote users should not rely on these characters for any type of professional advice. And also if the I guess they're able to scrub the content of the chats, if references to suicide or self harm come up, a pop up will direct users to a suicide prevention helpline, likely 988 or an online version of that. But some people argue that that's not enough because what ended up happening in the two cases that were cited by the APA chief executive, they cited two teenagers, a 14 year old boy and a 17 year old boy. The 14 year old boy. And I'm going to, and I probably should have said this at the top of the show, but I'll say it now that there's a kind of trigger warning here because I am going to be discussing suicide. The 14 year old boy in Florida died by suicide after interacting with a character claiming to be a licensed therapist. And the 17 year old boy in Texas had what they're calling, quote, high functioning autism. And after interacting with a chatbot that claimed to be a psychologist, there was a lot of kind of hostile and violent behavior that started to develop and that was particularly targeted towards his parents. So both of the boy's parents are now suing character AI because of what happened. That raised alarm bells for the APA as a whole. Like he was saying, basically, if this was a real therapist, they would have lost their license. But because it's an AI chat bot, there's no recourse. What do we do here? It's it's almost like it's a part of the design that these chat bots are going to mirror, mirror, mirror. So if you have a person saying I'm concerned about this, I'm worried that I might do this. Is there sort of a bug within the actual black box that is generative AI where they would say things like that sounds like a good idea, That's concerning. It's deeply concerning. And so how do we regulate something like this? That's an important question. It's one thing if a company is building an AI therapist and they're trying to market it. It's another thing if chat bots within a platform are popping up, whether it's the users themselves that are creating them or I don't even know how they're, if they're sort of like self creation within these platforms, how do we police that information? Are disclaimers enough, especially when we're talking about children on the platforms who may not understand the difference and honestly shouldn't be engaging in anything even remotely claiming to be therapy without consent of their parents if you make the companies.

Line 420:		Line 420:

	{{anchor\|news_item3}}		{{anchor\|news_item3}}

	=== Redefining Dyslexia <small>(43:31)</small> ===		=== Redefining Dyslexia <small>(43:31)</small> ===
	{{shownotes		{{shownotes

Mheguy: Page created (or rewritten) by transcription-bot. https://github.com/mheguy/transcription-bot

2025-03-02T00:21:00Z