Show HN: I made a cheap alternative to college-level math & physics tutoring
A new platform offering asynchronous, video-based math and physics explanations for about $10 per week is pitched as a cheaper alternative to traditional one‑on‑one tutoring, aiming to “democratize” access to college-level help. Commenters debate how well such a model can substitute for interactive tutoring and office hours, raise concerns about quality control and UX, and compare it to both AI tutors and existing resources like Khan Academy or MIT OpenCourseWare. Many highlight deeper structural issues in higher education—such as misaligned incentives, “learning debt,” and overcrowded or underused office hours—that create demand for new kinds of supplemental teaching tools.
Product concept and perceived value
- Service offers asynchronous, video-based explanations to math/physics questions for about $10/week, pitched as ~1/10 the cost of traditional tutoring.
- Many see it as a promising, more affordable way to “bring tutoring to the masses,” especially for college-level STEM.
- Several compare it to “video Stack Exchange” or “Codementor for math,” with videos reusable by many students.
Tutoring vs. one-on-one teaching
- Multiple tutors and professors argue that real tutoring is highly interactive: noticing hesitations, probing misunderstandings, tailoring problems, and iterating in real time.
- They doubt that watching someone else’s Q&A (or a one-shot video answer) can substitute for that individualized process, especially for weaker students.
- Others think it can still serve above-average or motivated students who already know what to ask and just need targeted explanations.
AI tutors vs. human explanations
- Thread is divided on LLMs as tutors:
- Some report excellent experiences with GPT‑3.5/4 or Claude, especially when models generate and run code to check math/physics.
- Others recount serious hallucinations and subtle math errors that are hard to detect unless you already understand the topic, warning strongly against relying on LLMs for learning.
- Consensus: AI is useful for hints, conceptual explanations, and debugging understanding, but not yet reliably authoritative, especially for advanced math.
University incentives, office hours, and “learning debt”
- Long subthread on how universities prioritize research and prestige over teaching, especially at elite schools and in CS.
- Mixed reports: some experienced packed office hours and restricted materials; others as TAs/professors struggled to get any students to show up despite extensive availability.
- Several note “learning debt”: once students fall behind, it’s very hard to catch up; systems rarely address accumulated gaps across years.
- Some suggest the service could help busy or working students who can’t attend in-person office hours.
UX, pricing clarity, and scalability
- Users flag mobile usability, confusing UI elements (3D carousel, “servers”), poor search, icon blocking, and unclear use of quotes.
- Multiple ask for clearer, upfront pricing rather than only “1/10th of tutoring.”
- Questions raised about economic scalability, content moderation, and quality control as more tutors join; ideas include ratings, flags, cross‑verification, and FAQ catalogs.