pgbovine

so exciting to see these ideas out in the world! i'm now imagining a Scratch-like playground for kids to explore end-user programming / AI in an accessible way, like some of the example apps you've shown

there's been a rising tide of academic HCI work in a similar space, wonder if there will be cross-pollination of ideas along these lines (many more papers i'm sure but some off the top of my head): https://arxiv.org/abs/2305.11473 https://arxiv.org/abs/2309.09128

also, while i have your attention here, since you wrote that related post on (not) vector db's ... what would you recommend for a newbie to get started with RAG? let's say i have a large collection of text files on my computer that i want to use for RAG. the options out there seem bewildering. is there something simple akin to Ollama for RAG?

very cool idea! i was also very interested in this problem during grad school ... prototyped an approach by hacking CPython, but the code (python 2.6? from 2010 era) has long bitrotted: https://pg.ucsd.edu/publications/IncPy-memoization-in-Python... https://pg.ucsd.edu/publications/IncPy-memoization-in-Python...

Your work is an inspiration as always!! My n00b question is: what do you think is currently the most practical path to running a reasonably-sized (doesn't have to be the biggest) LLM on a commodity linux server for hooking up to a hobby web app ... i.e., one without a fancy GPU. (Renting instances with GPUs on, say, Linode, is significantly more expensive than standard servers that host web apps.) Is this totally out of reach, or are approaches like yours (or others you know of) a feasible path forward?

Cool work! You and your team may be interested in these two recent CHI papers from Microsoft Research, both on very relevant topics to what you've been doing:

1) “What It Wants Me To Say”: Bridging the Abstraction Gap Between End-User Programmers and Code-Generating Large Language Models (https://arxiv.org/abs/2304.06597) -- they try to tackle a similar problem as what you described above

2) On the Design of AI-powered Code Assistants for Notebooks (https://arxiv.org/abs/2301.11178) - uses Mito as part of their case study

as the son of a university professor (mother) and an entrepreneur/MBA/businessperson (father), i totally second this observation. +1

ping me by email (see profile) if you want to brainstorm privately; my two cents is biased toward academia if you have an above-average setup (don't know until I see your CV!); startups come and go with each hype cycle.

i totally know how you feel since i have felt that way in the past as well :) here's something relevant: http://pgbovine.net/practical-reason-to-pursue-PhD.htm

my two cents: find something where the grants align with your interests, or find a brand-new faculty member who has startup money so isn't as bound by grants in the near term. and if things don't work out, simply leave and go to industry -- with a marketable skill set like what you get from being a CS major, there's no way for you to lose. go wherever has work that you like more.

totally agreed! things can change even from one year to another with the same advisor. for instance, i'm not the same advisor to my students that i was last year, or the year before that, or the year before that. (i've only been at this for 4 years, and each year is incredibly different from the prior one.) circumstances change, resources change, and constraints change.

you've inadvertently answered the question for me, to a first approximation :) i think that academia is a great launching point for a wide array of scholarly activities that the free market (i.e., industry) doesn't directly pay for: research, public policy, outreach, teaching, mentoring, industry collaborations, etc. i can work with whatever companies i want (even ones that are actively competing with each other at the moment!) and be seen as a "neutral" party; i can share knowledge via teaching and research again with a "neutral" voice without being seen as a spokesperson for a particular company or other special interest group. you're right, though -- there's a whole lot to the story. maybe someday i'll write something up!

thanks! that's a hard problem! i frame it in terms of critical path: http://pgbovine.net/critical-path.htm

if you can't get on someone's critical path, then you have to make it very easy for them to help you with very little time commitment. e.g.,: http://pgbovine.net/how-to-ask-for-help.htm

no way! (i'm glad i did it the first time, though)

thanks! AMA.

(no guarantees that i'll be able to answer via text, though; maybe i'll make a video later. been trying to minimize my computering time off-hours due to increasing wrist pains ... PSA: take care of your wrists, everyone!)

overnight successes are usually 10-15 years in the making; so much goes on before the public release

see here for US-oriented details: http://pgbovine.net/professor-job-titles.htm

Lecturers are part- or full-time teaching faculty (usually not tenure-track, but some schools like UC's have tenure-track Lecturer tracks) e.g., https://profiles.stanford.edu/48960

yep, this is often confusing for folks coming in from other countries: http://pgbovine.net/professor-job-titles.htm

oh noes, not that guy again! this was written over 4 years ago before i became a professor, so YMMV :) i haven't looked at this for years; this is a more updated (and complementary) take on some of these issues, how that I've advised ~25 research students so far: http://www.pgbovine.net/managing-me-as-your-advisor.htm

btw this may shed some light on your situation ... tl;dr unicorn jobs don't exist (for long) http://pgbovine.net/unicorn-jobs.htm

If you're cranking out papers, then you should have enough papers to finish soon; in that case, just finish your Ph.D. and move on. Tell your advisor that you want to graduate within the next year, and that you have the papers to do so. Like it or not, the credential can open up doors for you in the future in unexpected ways, and you're so close to it already. (If you were, say, a 2nd-year student and felt this way, then leaving would be a much more feasible option.)

creator of http://pythontutor.com/ here ... skulpt is great and i know one of its maintainers quite well. one use case that's important is using it to distribute self-contained zero-install tutorials onto computers such as school machines and those in less-well-resourced parts of the world without good Internet access. you can deliver an HTML/CSS/JS zip file on a USB stick or CD-ROM and get lessons up and running even without Internet access.

that said, though, one reason I haven't made the plunge to Skulpt and still choose to run Python server-side is that as students get more advanced, they will inevitably run into edge cases and other weird behavior where skulpt will not match CPython, and the error messages and other outputs they get will become inconsistent with expectations. that's why i still continue to run CPython server-side in a sandbox for http://pythontutor.com/

agreed! i cover this topic here: http://pgbovine.net/PG-Vlog-26-what-does-academia-mean.htm

relevant paper: https://homes.cs.washington.edu/~jfogarty/publications/works...

This is an example of research as illustrated by Figure 5 in that paper (known functionality, novel techniques)

What does "academia" mean to you? Are you an undergraduate student, a grad student, a postdoc, a research scientist, a professor (and if so, which level)? Relationships to software creation differ at each level.

If you're doing needfinding for a project idea to enhance learning, my suggestion is to focus on a single DOMAIN of learning first rather than trying to make something to help learning in general, since it's such a vague and broad problem. e.g., "how do you learn to do your taxes?" or "how do you learn basic home maintenance?"

does altruism count as vanity? people often feel good sharing information with others and helping others. perhaps that's still "vanity", but i don't think so.

to reinforce one's own understanding. you're motivated to get things right if you know that other people will be reading it, rather than, say, writing something in your own private notes.

This is something I've thought a lot about in the past. This short vlog episode of mine may give you some ideas: "For Aspiring Creators, Content Means Everything" http://pgbovine.net/PG-Vlog-9-content-means-everything.htm

I, too, wanted a Unicorn Job of this sort: http://pgbovine.net/unicorn-jobs.htm

After some trial and error bouncing around in industry/postdoc/etc., I've found that being a PI in a field where I have to raise relatively small amounts of money (compared to Big Science labs) to be the best balance for me at this point. I agree with Fomite that unless I magically get amazing at grant-writing, there's no way that I could afford to sustain full-time staff ... I don't envision myself going down this path since I want to minimize time spent writing grants and maximize my research/teaching/service time.

how could they need hundreds (or are they at thousands?) of software engineers?

supporting hundreds of millions of users (and probably tens of millions of paid users) is non-trivial