Audio Of The Week: Albert Wenger and Mike Masnick on Encryption
My partner Albert went on the Techdirt podcast last week and talked with Mike Masnick about his thoughts on encryption, secure elements, and other issues related to the ongoing debate about privacy and security.
Continuing to debate this topic is a good thing, because it’s difficult to see the whole picture. I agree with Albert this isn’t a black/white issue. The subtleties need to be cleared up and understood.What is clouding this debate is that we are intermingling encryption (the technical aspect) with freedom of speech, security, privacy, law order, rights of secrecy, terrorism, etc…Lots of subjects here.What I’m hearing Albert debating is that this about choices of trade-offs, and he is saying there are trade-offs for any side, and that he wants us to think a few steps ahead about the implications of these trade-offs, and not be short sighted and rushing in drawing conclusions too early.I would love to see Albert’s points laid out in a white paper format. I know he’s written and talked about it in different places, but putting it all in one place would be helpful to see the complete picture, especially the questions he asks and the analysis and thoughts he has already provided. This is a 5,000 word type of paper that could be written by a hired ghost writer. Publish it and let us debate it under more informed circumstances.
Maybe when he finishes his book
Will he have a chapter on reductionism? I hear what he’s saying about how reductionism isn’t helpful.Chomsky: Yeah. In fact, the reductionist approach has often been shown to be wrong.* http://www.theatlantic.com/…I’ll share slides on why & how we need to defeat Descartes later today. His is the philosophy at the root of the reductionist black&white, binary 0 or 1, “in the box” ways in which we end up thinking about nuanced issues like encryption, ethics, etc — when really we need to invent better frameworks.And it’s also at the root of why the machines can’t understand Nat Lang.
I understand USV’s position and agree with the principles of being able to access someone’s data if it’s to prevent and/or resolve crimes like terrorism and the harming of children.However, the mathematical laws of encryption (and indeed AI) lag the human laws of language and ethics. So putting in frameworks for the PRACTICE of access is challenging.I envision new forms of technical encryption (mathematical language integrated with Natural Language, genomics or biometrics).However, that is a 10+ year horizon and not something VCs would invest in because the research hasn’t yet even been done by the academic institutions.For now, we navigate the huge dissonance gap between what the mathematics of the machines can do in encryption and the complexity of diverse human considerations, context and ethics as best we can.
The other aspect of it is to do with how economics has defined “trade-offs”, the constituents of the agents in play and how we calculate for it within equations like Nash’s equilibrium and other game theory.One of the facets of mathematical economic equations is that they remove as much subjectivity as possible.Only “objective” measurements (interest rates, share price, yields, IRR, WACC, ROE, time, etc) are factored into the variables.Ergo “trade-off” equations can’t and don’t include the relative subjectivity of what’s ethical or not and the extent of how ethical.This is why it’s a philosophical framework AND a pragmatic tools problem at a seriously deeper, nuanced level because there are connected impacts across multiple sectors and technologies.
However, that is a 10+ year horizon and not something VCs would invest in because the research hasn’t yet even been done by the academic institutions.It’s not clear to me that Albert’s interest in this is for investment purposes. That might be justification in his mind for spending time on these issues, but from a traditional business point of view it’s actually taking your eye off the ball the way that I see it. Maybe maybe it might lead to something investment wise. But you’d have to compare it to the same time spent on something else a bit more concrete.  Ditto for the basic income arguments. On the other hand it’s better in my mind than leisure time watching sports as a hobby. Same type of thinking that parents use to justify time spent by their kids on esoteric pursuits. Or sending your kids overseas and saying that it’s a great experience that will help them further in life (when the idea is to make sure they get a job right out of college and perhaps the time and money might be better spent on something closer to the goal and not years in advance or for the purpose of “well rounding” a kid.
The encryption debate is also connected with the potential of Blockchain to do and manage escrow keys and USV is invested in Blockchain.So Albert’s passion for this topic is understandable.The objective of Blockchain is to swap out the old paradigm where the major techco’s control the keys & data to a paradigm which is Open Source, auditable, multi-culture and reliant on the many rather than a handful of big techco’s.
Yeah but I just don’t think that is what is driving this though.
Either way, it’s great USV is publicly having these philosophical debates and being proactive.
Why does this soundcloud imbed insist on making me go to soundcloud.com if I pause the recording? There is no way to continue to listen here (Firefox 45.0.1) and restart it unless I reload the page (and then attempt to scrub forward). Clicking anywhere just takes you to soundcloud (and even a bunch of techdirt material which I have no interest in I just want to listen to this recording.)…
.In the discussion about encryption and security there is a need to inject a bit of organizational discipline in order to compartmentalize the issues and solve them one compartment at a time. Otherwise we find ourselves trying to do fine work with a shotgun.An example is the issue of basic Constitutional rights which cannot be bargained away under some theory of “greater good.” The Constitution provides for a means to amend it and if it needs amending, then that is the proper methodology of attacking that compartment. There is a reason why the majorities necessary to amend the Constitution are what they are.Within our continental borders, the issue of law enforcement and what they should be entitled to do in pursuit of public safety and saving lives also contains an element of “obstruction of justice” whereby our laws provide that if anyone stands in the way of law enforcement’s pursuit of a criminal investigation, they too are guilty of a crime.The arguments of privacy v the legitimate duty of law enforcement to pursue criminals, terrorists, bad acts — rings a bit hollow when you begin to count the victims of the failure to be on top of this situation.In Brussels, as an example acknowledging that is not the US, one has to wonder as to what genius passed a law saying law enforcement could not tap phones or delve into Internet footprints while being forbidden to conduct surveillance or raids between the hours of 9:00 PM and 5:00 AM.They have paid a price for those policies. Ill-advised policies, no?Outside our continental borders, the CIA is, essentially, empowered to break any and all laws they want. The CIA can enlist other entities to assist them, like the NSA, and as long as they are operating outside the continental US, they can smash any laws they want.While this is a fairly broad swath and, clearly, there have to be some limits, the CIA exists to obtain “intel” by any means possible and in the digital age (with the assistance of the NSA) there is damn little they cannot wiggle their way into.Importantly, the CIA is absolutely forbidden to operate within the continental US other than providing counterintel, and by extension counterterrorist intel, to the FBI. I often wonder how well this is respected given the nature of the terror threat.While I do not trust our government, it is a better bet than say China.If broken into smaller edible pieces, this can all be solved.JLMwww.themusingsofthebigredca…
I often wonder how well this is respected given the nature of the terror threat.Kudos to them for being able to keep it hidden in order to protect us.  Also it’s obvious that the police stretch the limits of what they are legally allowed to do in order to get the job done.
While I do not trust our government, it is a better bet than say China.My dad came from a bad place and I never heard him say one negative thing about our government. Nixon? “They all cheat”. Nobody wants a boy scout running the country they want someone who is willing to test limits and get things done. This anti government shit really bothers me (you see this constantly on Hacker News). And now the people who are writing stories about moving to Canada or elsewhere because Trump might be elected. So absurd.
Agree, without Constitutional rights as the original framers intended, the answer won’t be good no matter what it is. That being said, agree that much of this debate isn’t black/white. We need to think about a dichotomy of choice. First step would be to have techies tell us what is possible. Then have Federalist focused Constitutional lawyers tell us what is within the bounds of the Bill of Rights. We will get a range of options, choose the one with the best probability of doing no harm.
On the issue of logs … ALL of the tech talks I’ve been to by techco’s from AirBnB to Spotify to HP have emphasized the importance of continuously logging EVERYTHING for purposes of instrumenting efficiency improvements and better algorithms to make recommendations for users.Are users aware of this? Probably not — other than checking the box in ToS.
Not logging things is a holdover from the time when disk space was expensive. Same as writing efficient code (elegant where not needed) where it actually doesn’t matter (because processors are so powerful now you can get a bit sloppy unless you are writing code for certain types of organizations where it does matter).  I have always (even when space was at a premium) been a proponent of logging anything and everything for diagnostic purposes and as a paper trail. It’s cheap and very often comes in very handy. It’s actually a bit amazing to me that this has to be pointed out to people it’s something (the paper trail back in the day) that has roots in legacy business practices.Note: This does not apply to all types of businesses obviously and I am a big proponent of speed just not elegant code and speed where it doesn’t actually matter and just is mental masturbation. Many times something that I write for my own purposes is viewed by someone who really knows what they are doing and they always make comments like “no don’t do it that way, the ‘right” way is this way”. Why? It works for me the way that I do it. Why do I care if it’s not the way they teach in CS class or a Perl one liner? Who cares? It gets the job done and prevents me from having to learn more than I need to for what I am trying to accomplish.
i wonder if US government agencies are systematically outsourcing their ‘needs’ to actors in overseas sovereign territories to evade legal oversight? i’d never heard of Cellebrite before now.
It would be fantastic if we could all video/audio of the week in a podcast player =)