Contents

Intimate Anonymity

Social media has pervaded our society. It is integral to how we communicate and interact. It has become imperative that one be active virtually as well as physically if one is to maintain a network of social familiars. In this post I will use a few examples of well established networks: Facebook; twitter; 4chan; and reddit, to illustrate two very different types of social network. This list is by no means exhaustive, but they are the colloquial tips of the iceberg.

Social Networks

There are two broad categories in which social networks operate, personal (Facebook, Twitter) and anonymous (4chan). I am a strong advocate for anonymous social networks. Anonymity is a shield behind which people are free to express unpopular opinions, project feelings, and ultimately communicate free from the unspoken rules of society. The major drawback to pure anonymity is the very thing that makes it unique, no individual can know with whom they are communicating. This is where personal networks operate. In a personal network one can cooperate based on shared in person experiences.

The Idea

Neither of these established networks should change. Both have their pros and cons. However, I posit that the dichotomy of personal vs anonymous should be eliminated in favor of a more linear scale. A network that would provide anonymity only after verifying a persons existing relationship. This intermediate form of social networking is something I call “intimate anonymity”. There exist pseudo-anonymous networks, that is, networks where a persona or identity is established, but not connected to a physical person (reddit). This is a rough, difficult bridge between anonymity and personal and is ultimately an obfuscated form of personal networking.

Social Structure

Below are the three relationships by which any two individuals are related. These are presented as a cascading hierarchy. The "social contract" referenced is the mutual understanding from both parties of the status of their relationship (IE: Both parties must agree).
Individual no IRL interaction no social contract
Acquaintance little to moderate IRL interaction no social contract
Friend frequent interaction social contract required
Using these we can identify on what level each form of network allows anonymity and in which direction (an arrow indicates that actor cannot see with whom they are interacting):

Anonymity (Both individual and poster)
	Individual <-anonymity-> Poster
Intimate Anonymity (Only friend)
	Friend -anonymity-> Poster
Personal (No anonymity)
	Individual -anonymity- Poster

As an analogy, imagine a room full of individuals:

Anonymity is a room full of unidentifiable people selected at random.
Intimate Anonymity is a room full of unidentifiable friends
Personal is a room full of identifiable friends.
In this situation individuals can rely on shared experience (IRL) while also speaking freely from a place of anonymity. So how can such a system be implemented? This is where my concept begins to tread shaky ground. Still, I have a few ideas. First, terminology.

How does it work?


Userspace
	A network of individuals that is curated by a poster
System
	The network itself.
	Any actions directed here imply some form of user experience.
	These actions are only initiated by a poster, and the results only affect the poster's userspace.
Individual
	A person using the system to interact with a poster.
	The individual is allowed to know the identity of the poster.
	There can be many individuals in one userspace.
Poster
	A person using the system to interact with individuals.
	The poster is not allowed to know the identity of these individuals.
	There is only one poster in a userspace.
User Key
	A unique key, generated by the system.
	The user key can only be created by a poster for their user space.
	This key can then be given to a single person, allowing them access to the poster's user space.
Using this terminology, the following is a work flow in which a poster adds an individual to their userspace:

Poster requests a user key from the system
The system generates an associated private and public key
The system gives the public key to the poster
The poster gives the public key to the intended individual
The individual adds the public key to their profile

When an individual attempts to enter a userspace, their key is checked by the system and they are only allowed if the associated private key exists and is associated with the space they are attempting to access.

Final Thoughts

This will probably require a more complex structure, but this is still just an idea floating in the back of my mind! I'd welcome constructive criticism. It would be nice to know of any blatant flaws in this design. There would be other interactions with the poster and the system such as banning an individual and probably many more, but that's down the road. Until this exists, I'll have to continue navigating IRL interactions and wondering what is and isn't appropriate to comment on.

Intel or AMD?

This question is inevitable, predictable, even expected to follow the innocent statement that one has built a new computer. The resulting conversation, however, is anything but innocent and might be more appropriately housed in the Colosseum than a computer shop. My choice to move to strictly Intel builds is fueled by the following: Financial Intel spent 10.6 billion on research and development in 2013 (Page 11). Compare that to AMD's 5.2 billion net revenue (Page 39) for the same year. Intel literally spends twice what AMD makes in a year on R&D alone. It's no wonder they are the bleeding edge in regards to performance. Intel's Net revenue in 2013 was 52.7 billion (Page 27). Such financial stability most assuredly affords them the ability to innovate in areas that, if unprofitable, would ruin other companies. AMD's net revenue has been hovering around 5.3 billion for the past five years. You will notice that in 2010 their net profit was 6.4 billion (Page 39), and in 2011 it was 6.5 billion (Page 39). So what did they do right those two years? In 2010 AMD posted a 325 million gain from the deconsolidation of GlobalFoundries Inc. for the quarter ending March 27 (Page 39). 2011 was a leap year, awarding AMD (and everyone else operating with our concept of time, and in our observable dimension) an extra seven days of income. This isn't to say that I do not feel bad for AMD, who has long long suffered at the hands of Intel, and has attempted to fight back. This shows that the consumer will go where the grass is greenest. What about AMD's innovation? After all, in the eighties the 8088 and 80286 processors could be clocked past the Intel counterparts. The issue arises in that AMD was building CPUs based on designs given to them by Intel, down to the actual numbers used to identify a certain architecture. The following is an excerpt from an interview with Jerry Sanders (Chairman & CEO of AMD) in 2000. "When we introduced our products, instead of calling it some other number with a cross-reference chart, I called it AM for AMD, 741, AM747, AM723. In the case of National, AM, let’s see what did they have… AM108, AM101A, so everybody knew exactly what those numbers were because National and Fairchild had been spending fortunes promoting them. They were designed everywhere.
Now I have to tell you, everybody wrote me letters saying that I couldn’t do this, that they were going to sue me. But it turns out they couldn’t, because they had no legitimacy on claiming a copyright of merely an alphanumeric. Of course subsequently, decades later, Intel sued me when we came out with the AM386. They said, “You can’t do that." But we won in court. Of course we could do that. That’s why since we had the AMAD88 and AMAD86, and the AM186 and the AM286, the AM386, and the AM486, Intel said, “Enough of this," and came out with Pentium. There was a name which was now a unique name and copyrightable."
Full Text. Jerry Sanders has also been quoted as saying “I understand what you're thinking. I spend more than I make. I always have spent more than I make." in regards to his house (for which he outbid Madonna). AMD was founded by a man with poor financial management, and it has plagued the company since. If the company can find itself in a stable financial situation, they would have the footing necessary to compete on the same level as Intel. Performance Intel consistently outperforms AMD. CPU-Benchmark-dot-com's PassMark list is dominated by Intel CPUs, with the first AMD (the AMD FX-9590 Eight-Core) appearing thirty-sixth with a score of 10,290. The Intel Core i7-4790K sits at twenty-eighth and a score of 11,436 for a comparable price. Let us, for the sake of argument (and ease, as the hard work has been done already), compare the Intel Core i7 3770K and the AMD FX 8350. The i7 has: much lower idle power consumption (75w vs 92w); newer manufacturing process (22 nms vs 32 nms (Which turns out is really small)); better performance per watt (8.28 pt/w vs 5.21 pt/w); and lower peak power consumption (128.3w vs 182.21w) among others. For an extremely detailed look at how Intel vs AMD bottlenecks ATI GPUs, read this). The Intel build performed marginally better than the AMD build, and only cost $180 more, a fraction of the almost $2,000 price tag on either build. It should be noted that ATI GPUs require much more from the CPU than their Nvidia counterparts. This anandtech article only solidifies what was already expected. Intel CPUs outperform AMD on particle movement, multi and single threaded bechmarks and others. While the AMD excelled in areas that had been optimized to utilize more than four cores (which very few are). With all that being said, my next build will be Intel. If AMD is pulling ahead it is by brutish clock cycles and lower price tag. But Intel will outperform with surgical precision in all areas except those specifically optimized for >4 core processing.

Blogging is hard

For me, specifically. Each post here is the realization of successfully selecting a topic, digesting enough information to provide a summary, and maintaining focus on that one subject long enough to write the post. Selecting a topic, I expect, will be the least painful of these three. Digesting information isn't so hard, once that information has been found. The last, maintaining focus, may prove to be troublesome. Answering one question typically raises two more, and I am impatient. I must know! I foresee two possible writing styles emerging from this absurd experiment. In the first, the removal of all boundaries allows for a relaxed, almost a piacere piece. In the second, the UN-inhibited free flow of thoughts into text results in a bloated and in-decipherable amalgamation of reflections. In either case, I hope to learn something about myself.