anybody want to test an ai tool that helps kids build oral reading fluency? it's recommended to build oral reading fluency if: - the student has mastered phonics - the student can read aloud, albeit slowly - reading aloud requires a lot of effort

when i’m not building timeback or tweeting, im talking to students about rubik’s cubes

can someone explain to me what in-the-weeds technical constraints prevent 100,000x learning rate at same quality

i don’t think it’s really that hard to make kids love learning kids are born curious so just don’t do whatever schools do to make kids hate learning

my understanding is: phonics is good, phonemic awareness is good. the RCT suggests that *advanced* phonemic awareness is unnecessary in other words, it’s useful to know blending and segmenting when it comes to phonemes/sounds. not so useful to know advanced manipulations like

alpha school is the purest learning science laboratory in the world the software is tested and only goes to students if it gets better and faster learning outcomes since it’s code, the experiments are perfectly replicable

i think the past 2 days is the longest i've gone without tweeting in a long time i've been super busy building and the plan is to pilot oral reading fluency app with students in december at the latest if it drives learning outcomes better & faster than what students are

"there are some psychos out there that think that's a good time" can confirm. as a psycho, it was the best time of my life

duolingo wants kids to spend 24x longer learning. i want kids to learn 24x faster duolingo's incentives are off. they want you to never churn from the app. timeback is incentivized to give kids time back for other school activities from interview with duolingo founder:

now imagine this, but with a better ui and even better learning science that's what we're building

lol a student tried to change their XP in developer tools and timeback flagged it as cheating

i have not traditionally been a huge fan of reading the classics, but here's my best attempt to make the case for it: Lindy effect — books that were popular for a long time are likely to remain popular for a long time it allows you to see similarities and differences in culture

not only is alpha school the first school to do academics with ai, it might be the first company to manage employees with ai i think this will be the norm in 5 years

imagine a reading app: - spaced repetition across every concept needed for reading. phonics, vocab, spelling, grammar - engaging and varied texts to build fluency and stamina. include support for logging physical books am i missing anything needed to make it an optimal reading

timeback is more than a learning app. it's an operating system timeback solves all of these questions: - how do you measure student progress and spot struggles? - how do you decide which apps a student should use? - how do you reward students?

1 year later, nano banana has solved this

the Simple View is Reading states that Reading Comprehension = Listening Comprehension * Word Decoding a longitudinal study found the model was quite accurate for younger kids—explaining 96% of variance in reading comprehension! further, since “listening comprehension” is so

ash is out here revolutionizing education for adults i know because i learned from him and i wouldn't be where i am without his help

the best thing about ai teachers—they’re easily coachable suppose you tell a human teacher about education best practices. they might follow it, they might not with an ai, i run an evaluation every time i update the prompt or configuration. so i can see in seconds if it’s not

hear me out: school that uses timeback for academics but you just do academics all day 20x learning