@BylinesEast Current LLMs can produce a decently written article about virtually anything, certainly; but none of the current generation can produce a well sourced article at all.
They can invent sources, sure; but by the time you've sorted out which sources are wholly fictitious and which are real but don't say what they claim, you've done as much research as you'd have needed to do to write the article in the first place.
AI News: Could not possibly do worse.
Let me prove you wrong.
Using #AI effectivelly is a learned skill.
Give me a subject you are familiar with and I likely am not.
I like a challenge.
@n_dimension @BylinesEast Sure. Should we start with the impact of #LLM use on human cognitive function?
(Please don't confuse LLMs with AI)
I am familiar with this paper.
There are numerous issues with it.
No. 1 being, it is not peer reviewed.
Its a preprint.
No. 2 The sample cohort is minimal and the data obfuscated/not provided. They use language like "linguistic underperformance" without providing values.
Is it 0.3% or 30%?
No. 3 Speaking of language, its using biased language "cognitive debt". Which is not what scientific inquiry is about.
There are multiple red flags with the preprint.
Also, it's like saying "When using exoskeleton for physical work, your muscles work less".
Yes. That's the point of using the assist tool.
Circular reasoning.
My own take on "loss of #cognition" using #ai is here;
https://infosec.exchange/@n_dimension/114657347248606224
Attached: 1 image Critical reasoning vs Cognitive Delegation Old School Focus: Building internal cognitive capabilities and managing cognitive load independently. Cognitive Delegation Focus: Orchestrating distributed cognitive systems while maintaining quality control over AI-augmented processes. We can still go for a jog or go hunt our own deer, but for reaching the stars we, the Apes do what Apes do best: Use tools to build on our cognitive abilities. AI is a tool. 3/3 #AI #Education #FutureOfEducation #AIinEducation #LLM #ChatGPT #Claude #EdAI #CriticalThinking #CognitiveScience #Metacognition #HigherOrderThinking #Reasoning #Vygotsky #Hutchins #Sweller #LearningScience #EducationalPsychology #SocialLearning #TechforGood #EticalAI #AILiteracy #PromptEngineering #AISkills #DigitalLiteracy #FutureSkills #LRM #AIResearch #AILimitations #SystemsThinking #AIEvaluation #MentalModels #LifelongLearning #AIEthics #HumanCenteredAI #DigitalTransformation #AIRegulation #ResponsibleAI #Philosophy
It looks like we both failed successfully.
I love these challenges as I always learn something.
My failure: 🚫 Use of LLMs can impact your cognition, though Longitudinal data is lacking and there is a performance increase at the time of use.
Your failure: 🚫 LLM can create an accurate and well sourced article (if you know how to use the model).
Further, if you accept the veracity of the article, you falsify your own premise (which was based on flawed data) /chefs kiss 😘
Thanks for playing.
#Ai #LLM #cognition #criticalreasoning #psychology #cognitivescience
https://claude.ai/public/artifacts/5d67ba34-d702-48bd-8b41-706fb8b2a9af
East Anglia Bylines
Simon Brooke
Wulfy—Speaker to the machines