is AI the future of learning?
What do you think the personalized AI tutor everybody gets hyped about will actually look like or do? Do we have the capability already? is it even possible?
Learning from AI requires certain skills. You need to ask the right kind of questions. You should use AI together with other sources of knowledge, not as a replacement for them. For example, you can read a book, and when you have an additional question, you can ask the AI. Or you can ask the AI first to provide you a high-level summary of a few books, so that you can choose the right one to read. You can try to solve a problem on your own, and afterwards ask the AI how it would do it. You can ask AI to comment on your solution, or to modify it.
It is not possible to learn while remaining completely passive. Ultimately, learning is an activity of your brain in response to the inputs you get; if you do no activity, there is no learning. But AI can give you the inputs you want, quickly. (Much faster than a search engine, which was a huge improvement over searching various catalogs, which were a huge improvement over clicking random links and hoping to arrive at something relevant.)
Maybe one day there will also be an AI tutor for lazy people, where you won't even have to ask the right questions, because it will be programmed to think for you. (It might contain a loop of asking itself what is the right question you should ask at this moment, and then giving you both the question and the answer.) I don't see a reason why such thing couldn't exist in principle, but I also expect the first few implementations to contain some big flaws.
A good teacher or tutor should know when to let the student struggle a little bit so they can discover the answer on their own to create some self reliance and confidence.
They might not be a perfect model in all senses but for people who know how to use them well imo they are spot on.
The only caveat I see is that the person should at the very least understand how these ai tools work to leverage them properly, for example: if you have no idea where to start coding, you can get a roadmap and start working but knowing how chatgpt generates such a roadmap for you might help in optimizing prompts to get the best out of it.
hope this makes sense
This week on the Odd Lots podcast [0] there was an excellent interview with the founder of Duolingo who talks about how they use AI to personalize the learning for each user.
But even before that, we had adaptive tests. Not AI, but on the path to adaptive learning systems.
[0] - https://www.bloomberg.com/news/articles/2024-05-02/luis-von-...
Cons: - Jailbreaks enabling suicide ideation (huge issue in APAC) - Hallucinations - Dehumanisation and social isolation. Already the biggest issue we see from people entering the workforce is inability to navigate the workplace socially and to negotiate around conflict. LLMs are conflict avoiders.
- In the end this is commoditization. Poor people get AI tutors, rich kids get face time. PE splits the difference
It was a dream of mine for ages to bring to life the courses I run in an online manner, and use it as a vehicle to help newly-qualified teachers. Without AI, it probably wouldn't have been possible.
For those who want to break out of the lot they've been dealt in life, AI could be a godsend when it comes to learning.
He had one routine where he said it's fun to teach wrong...
See, what you do, you have a three-year-old kid and you want to play a dirty trick on them, whenever you're around them, you talk wrong. So now it's like his first day in school and he raises his hand. "May I mambo dogface to the banana patch?"
A human does this through building a relationship and understanding how their pupil learns. I don’t know how an AI model will do it.
LLM as a device fits so many use cases.
For a long time we have had the capability to use student learning record data to recommend optimal questions to maximise student grades in a given subject(adaptive learning); we've known the best exercises and methods to apply this with, like quizzes/active recall (see Roediger & Karpicke, 2006) and increasingly, made interfaces that keep people in the zone.
The problem is, such systems have 3 challenges compared to a human experienced tutor/small group teacher:* - 1. It's not abstract. It can't identify misconceptions in your wrong answers like a human would and refocus new activities on these gaps. Doing this is very effective. I believe this is possible in the future, hence why I developed blurt mode, a learning exercise on my website Revision.ai which asks students to blurt content they have learned on a specific subject, rates their coverage and accuracy and most important identifies, clarifies and stores for future exercise content their knowledge gaps, by using AI tech to identify misconceptions in a way we could not do 4 years ago. - 2. An app will never truly check up on you socially or with social expectations of appearing in real life like a tutor; it's never as urgent. Staying engaged for 6-7 hours rarely happens either unlike real life with it's physical location constraints (which also have a role in memory formation and retrieval) - 3. The perfect app needs millions of student learning records that account for your specific experience and knowledge. This is not possible. While humans cannot do this either, intuition of teachers is lost with apps.
*Small groups/tutors can provide customised, targeted and long-term feedback which you feel a sense of accountability for responding to according to the research from Bloom, and is largely responsible for why wealth correlates with education with higher attenuation at the top quartile than the differences between the bottom three quartiles - paying for private tutoring really is a cheat to achieving more on grades (and therefore more opportunities), although whether it improves intelligence, is debatable(nobody agrees what intelligence is really, beyond some early G factor work that personally I do not see as that valid).
I think you should be careful using the word "personalized" too - one, it is a popsci buzz word, and two, and consider "adaptive" instead - they have different meanings in the psychology of learning and like I address above, the personal aspect are mostly confined to humans.
"Never send a human to do a machine's job"
A better alternative:
Apply some AI!
Actual Intelligence!
go to the library and read some books.
get a masters degree by reading some books.
In fact just read books.
The lost art of reading books.
The lost art of being human.
Instead id like an LLM to point me to relevant hard fact written by a human on a subject, be that text book or some other long form content.
LLMs spit out crap constantly.