pkcsecurity

I think the problem with vector search is that it’s unbounded - is the 3rd result relevant at all? What about the 4th?

On top of that, I believe it’s clear with how Cursor and Claude converged on regex search after using vectors that asking LLMs to come up with similar terms to regex is more effective than searching for the user’s original term over a vector db.

It’s not just “seen as a bad deal by the company” - a business can’t give 100% of the value back to the employee because it has this pesky thing called “profits” it has to worry about :)

This is a pretty interesting approach.

We've been doing a similar pattern already, because we have our web app iframed and communicating with outlook (via Office.JS) and gmail (via InboxSDK.JS) in order to read changes to the To: field and insert contents into the compose page based on that:

* stimulus encapsulates the javascript logic that "bridges" to Outlook and Gmail, including initializing the To: field listener * stimulus actions can trigger the "bridge" as well

It's worked out very well - though I really hate dealing with native stuff in general - especially compatibility nightmares with older versions of Outlook. Probably less of a problem with iOS/Android.

Comment on the timing of this IPO: for a company to IPO "unfavorably" compared to previous valuations likely means ARM's hand was forced by timing considerations, unless they are running out of cash, which I don't think is the case.

My interpretation of this is that their investors suspect that whatever boost ARM is getting from AI optimism will likely peak soon, so the timing has to be now.

I know that's not fully rational because they're mostly unrelated, but certainly the optimism because of AI has to be good for them. Curious to see if this "signal" plays out - investors tend to be pretty savvy about timing.

What counts as “remote access”? Another device authenticated to Wi-Fi? Another device anywhere on the internet, with knowledge of the device ID? Another device anywhere on the internet with knowledge of email address?

These are vastly different criticality levels.

All the talk of IOCtl and assembly/bytes in the in the ButDefender report implies “another device on the Wi-Fi”, but I know wyze cams can be viewed over-the-Internet, ostensibly proxied via Wyze’s own servers, so maybe not?

I think part of that is the architecture and environment of the museum. They spend tons of attention to the high ceilings, sense of awe (a bit like a cathedral), and that can make all the difference compared to viewing something in your bedroom in the dark.

The New Yorker is great too, if you’re willing to pay a few bucks a month.

Moxie makes so many good critiques (some are so subtle, it might be worth a second read). I got the sense he’s trying very hard to be even handed and constructive about a situation he feels pretty badly about, but his true feelings are bleeding through in some of the side points / parentheticals.

One point that I disagree with is his almost axiomatic premise that decentralization is an inherent good and the implication that the Internet went wrong because it failed to stay decentralized. To hint at great cryptography as the solution, as he does im his conclusion, is baked deep in his bones as an amazing cryptographer, but I think he’s prescribing the wrong cure. The problems with the Internet are fundamentally not about decentralization - they’re about trust. It’s a people problem, not a technology problem. Because of this, cryptography (I do not mean crypto) simply cannot be the answer - even the best cryptography is, like a great legal system, only capable of dramatically reducing the overhead costs and risk of operating in a given environment. When it comes to what great cryptography can achieve, I think HTTPS and maybe some E2E stuff that’s happening with Signal is as good as it can get (interestingly, HTTPS is good in large part thanks to Moxie) - it cannot bring us back to some golden Internet age.

Watch some of the games and it might change your opinion. The thesis that "AI in Starcraft will only win via improved mechanics" is false - the AI was making some fascinating decisions / fundamentally different meta strategies.