work

All things considered, this was a pretty good work week. The worst part was realizing I have a lot of work to do for a DSO deadline coming up and not much time to do it. The deadline itself isn't till mid-August, but next week, I'm going on a two week vacation during which I expect to accomplish approximately zero work. That means I essentially have a week and change to write the rest of the report. This is probably doable, but it almost certainly will cut into my relaxing and fun time. Anyway, aside from meetings, this was my main focus this week, and I don't have much to say about it. I was mostly just understanding Attribute-based Encryption schemes well enough to explain it to someone unfamiliar. I don't like that it takes away from some research I want to do, but it definitely is worth it (and useful to do a deep dive into really interesting papers). I'm also very fortunate to have great collaborators who brought great and productive ideas this week even though I wasn't able to work on my own much. Hopefully some of these pan out and I have more to say in the future :-).

Oh I almost forgot! My recording with the Zero Knowledge Podcast went well as far as I could tell! It was really fun talking with Anna and Kobi about PIR (and a shout-out to ORAM at the end). I definitely need more practice talking about my work, so I was grateful to have the chance to do so with them. Feel free to check it out when it drops if you want to hear my broad thoughts about PIR and what's exciting.

non-work

Life outside of work was a alright. I feel pretty trapped by the wildfire smoke that's covering the area right now. I ended up just walking instead of running most days, since I read that it's better to not breathe as hard in the smoke. Walking is still good, but it throws off my routine quite a bit. I also stayed home more than I would've liked to as well since it was so bad out (and maybe also to watch some of the world cup games :-)).

It was ok being cooped though. I finally beat A10 on Slay the Spire 2 in multiplayer with some of my friends! I've also been revisiting Outer Wilds which is a really hard game to get back into. There are so many lore bits that I either didn't read carefully before or just forgot. Despite this, I hope I can finish it before I head off to Europe, since it was recommended by my friend who's getting married.

This coming week I am on track to wrap up The Mind is Flat which has crystalized a lot of the thoughts I had about consciousness. Just listening to the book and considering the theory of mind he puts forward has resonated with my experience at the time of reading as well as so many past experiences. I'm still not totally settled on the "hard problem" of consciousness, but I'm increasing leaning toward pragmatic/illusionist views that the problem itself is a pseudo-problem.

questions

• Does anyone care about PIR with preshared randomness that's independent of the database/query index?
  • This is a common approach in MPC that's used to speed things up, but it seems pretty silly since PIR usually has many clients and a known database
• Are "zero-knowledge" ideas going to come to PIR soon?
  • Kobi mentioned on the podcast that he thinks people will use ZK stuff in PIR, but I haven't seen this in many PIR applications I know of.
  • I know there are some generic authenticity things one could do with ZK, but I don't know what would be interesting
• Are there array PIR -> keyword PIR compilers that support efficient updates?
  • I did a little bit of research on this before the show, but I think I only found compilers that required re-encoding the database when there was an update.
  • I'm sure there must be something with something like an adaptively correct cuckoo hash or something, right?

Thanks for reading to the end! Here's a picture of Timmy with the cabbage guy and the wildfire smoke over NYC.