Jeff Coleman | Jeff.eth

11.8K posts

Jeff Coleman | Jeff.eth banner
Jeff Coleman | Jeff.eth

Jeff Coleman | Jeff.eth

@technocrypto

better worlds before bags @technocrypto on Farcaster @[email protected]

参加日 Kasım 2014
341 フォロー中9.6K フォロワー
Jeff Coleman | Jeff.eth
Jeff Coleman | Jeff.eth@technocrypto·
@jachiam0 @davidad @nlpnyc And then you still have to have some type of consideration for the downstream consequences of the microstates themselves, which a human would implicitly at least *try* to optimize over en route, but which may tend towards "undesirable side effects" in an unlike optimizer.
English
1
0
5
195
Jeff Coleman | Jeff.eth
Jeff Coleman | Jeff.eth@technocrypto·
@jachiam0 @davidad @nlpnyc There is also the concern that in practice the most useful tasks may be those where we can't define or constrain the microstates very tightly, so that even the lower K-complexity goals we *could* define won't be very useful at stopping people from just pursuing others instead.
English
1
0
0
714
Joshua Achiam
Joshua Achiam@jachiam0·
"not all goals lead to instrumental drives" is obviously true, but it is also heresy of the highest order in x-risk world, so people will fight you to the death on it without actually stopping to wonder if it is correct
English
5
0
4
4.5K
Jeff Coleman | Jeff.eth がリツイート
isaac Samuel
isaac Samuel@rhaplord·
short thread on when Romans and Africans from Kush visited each other 🧵 Marble bust of an ambassador from Kush, found at villa Albani (Italy), dated to the Flavian period (69-96AD) Roman visitor’s inscription, found at the Musawwarat al-Sufra temple (Sudan), undated #randomxt
isaac Samuel tweet mediaisaac Samuel tweet media
English
6
88
362
49K
Jeff Coleman | Jeff.eth
Jeff Coleman | Jeff.eth@technocrypto·
@Thomas_Gondwe_ Another interesting user model to think about is where there exists a smartphone with no data, and it can authorize transactions by outputting human readable strings which are then sent by SMS to perform authenticated actions.
English
0
0
2
47
Jeff Coleman | Jeff.eth
Jeff Coleman | Jeff.eth@technocrypto·
@Thomas_Gondwe_ Yes I've been thinking a lot about a "one smartphone per table banking group" or "one smartphone per cluster of community banking group" models.
English
1
0
3
49
Jeff Coleman | Jeff.eth
Jeff Coleman | Jeff.eth@technocrypto·
Been chatting a lot with this team about how they can integrate this idea into #ensdomains rather than needing to build a whole separate system! One thing I saw over and over on my recent 🇪🇹🇿🇲🇰🇪🇬🇭 trip was how mobile-first these ecosystems and economies are so I can...
English
5
10
42
9.3K
Jeff Coleman | Jeff.eth
Jeff Coleman | Jeff.eth@technocrypto·
@strwigod ...sense to do account abstraction stuff where you only do full cryptographic setup if they want to deposit larger amounts though, or stuff like that. Spread out the UX barriers/hurdles.
English
0
1
1
34
Jeff Coleman | Jeff.eth
Jeff Coleman | Jeff.eth@technocrypto·
@strwigod I'm definitely bullish on the idea of phone numbers as *UX* tools. But the thing to do under the hood if you want to turn them into actual protections is map them to cryptographic identities of some kind and then do social recovery on that, *most* of the time. Could also make...
English
3
0
0
36
Jeff Coleman | Jeff.eth
Jeff Coleman | Jeff.eth@technocrypto·
@strwigod Maybe I misunderstand "directly" but how could this ever achieve better properties than recovery schemes based on the federation of SMS verifiers themselves? Do you mean like texting shares to people?
English
1
0
0
190
Jeff Coleman | Jeff.eth
Jeff Coleman | Jeff.eth@technocrypto·
...audit the reliability of SMS verification providers or something? Very curious to hear who else has been working on this problem and what ideas they have come up with, and I'm sure the PNS team would be as well. Please reply here if you know of them! I'm eager to learn more.
English
5
1
12
2.6K
Jeff Coleman | Jeff.eth
Jeff Coleman | Jeff.eth@technocrypto·
...will be deciding how to determine when an SMS verification service is considered "authoritative" without making the whole architecture depend on one or a few of them. Maybe a peer to peer trust rating system where we text "free money codes" to randomly chosen addresses to...
English
1
3
11
2.6K
Jeff Coleman | Jeff.eth
Jeff Coleman | Jeff.eth@technocrypto·
@2plus2make5 @m0k1d For agents with sufficiently advanced and legible decision algorithms there exists an equilibrium with lower commute time, for example.
English
0
0
1
78
Jeff Coleman | Jeff.eth
Jeff Coleman | Jeff.eth@technocrypto·
@2plus2make5 @m0k1d They may get into it later in the course (and I realize this is ten years ago) but worth noting that the CDT dominant strategy is far from the weakest behavioral assumption. There are huge assumptions there about player decision algorithms and mutual (il)legibility of them.
English
1
0
1
127
Emma Pierson
Emma Pierson@2plus2make5·
man, I started watching Tim Roughgarden's lectures because I wanted to learn a bit of algorithmic game theory (youtube.com/watch?v=TM_QFm…) but I may keep watching just to learn how to teach. I'm not sure I've ever seen a stronger opening to a class.
YouTube video
YouTube
English
14
79
792
123.9K
Jeff Coleman | Jeff.eth
Jeff Coleman | Jeff.eth@technocrypto·
@BigGirlsRise @GBMauction Best of luck to your project! I hope you're successful in turning this into a community, foundation, and so on which really moves the needle on these issues ❤️
English
1
1
6
232