work

Thankfully, there's little to say work-wise. I pretty much took the whole week as a true vacation. I had a handful of emails that required more-or-less immediate attention that I responded too -- mostly just Eurocrypt reviews and a few meetings that I had to cancel. I'm excited to come back to work in the next week, refreshed, though. I've had a few research ideas I want to think about more, some teaching ideas, and some logistics I need to take care of too. Always grateful to have the flexibility to shift my workaround.

I guess while I was on vacation, Vitalik Buterin posted about Plinko on his blog! Very cool to see the work get some attention and hopefully an application for Ethereum.

non-work

Outside of work, on the other hand, I have had a very busy week! My partner came to visit Hoboken, and then we went to visit my family in South Carolina for Thanksgiving (where I'm writing this from). We managed to pack in a lot before we left Hoboken, though. We made pottery, saw Wicked: For Good, ate a ton of vegan dim sum, walked around Central Park, and more! It was so nice to have her back in the area in person and to enjoy the NYC/NJ area with her for the first time since September. Since we came to South Carolina, we also had a ton of fun (and more good food) with my family.

I also finished reading Four Thousand Weeks! There's not much to say about it beyond what I said last week. It was a good reminder to prioritize what's important, which is I think all it set out to do. Since then, I've started reading Mere Christianity by C.S. Lewis. This is one I've been recommended and heard a lot of people reading, but I've been putting it off because I'm not too convinced by most Christian apologetic arguments. I'm not finished with it, though, so I'll still try to reserve judgment.

Also, I've resumed reading Sara E. Igo's The Known Citizen in parallel. I shouldn't have put it down before, because it's really fun to understand how the concept has developed over time. Still have a long way to go through it though.

questions

• Is it possible to build verifiable PIR with client-side preprocessing that rely only on one-way functions?

  • I may have asked this before, but a lot of vPIR approaches require expensive computation that would be great to remove

  • If this does exist, then it would also serve as pretty good motivation for client-side preprocessing and might lead to it getting deployed more

Thanks for reading to the end! Here are a few pictures, since it's been such a crazy week.