I'm mostly concerned about government debt and bond yields. When governments start having a harder time bailing out the larger economy this can cause all sorts of problems, england atm is a prime example. If many governments start having this problem, FX contagion can start playing havoc with industry as no one knows who to export or import from or to.
The promise of crypto was that someone found a way to make a non-duplicable, tradeable token. First by proof of work and then by proof stake.
The problem is that they forgot the 1 against all adversarial nature of internet security. A trillion dollar empire set against one way of securing assets is in essence a trillion dollar bounty on breaking your security regime. And what was found? That the system was complex enough that most security "hacks" were mostly social engineering with a sprinkling of insecure gateways and passwords. Neither the users of wallets or the engineers of exchanges put in enough time and effort to secure their holdings in relation to the dollar values they held. Wild swings in the stability of coin values created arbitrage opportunities which made some users exceedingly rich, prompting more dumb money to enter the system. That's a bug, not a feature.
I still think there's value for crypto to evade censorship and manipulation of populations via legislation of the money supply. But that requires a whole slew of low level techs to maintain and verify transactions. At some level that may make it as onerous and bureaucratic as our current financial systems.
I just hope that this doesn't cause financial contagion in the traditional sectors.
I've seen more and more advertisements for Mastadon on HN, probably because of the whole Twitter debacle. Does anyone use the service? I mean, enough people so that the average person would know what it is? From my experience with Mastadon it took me a while to even learn how the service was used, and all I remember was thinking "wow that's too complicated to use or get Joe Average User to sign up for". Without ease of use a critical mass of users won't be possible so it won't ever be anything other than a niche product.
Someone else posted a link to their for profit search engine. I find having to login to use the product a bit disturbing. What if I don't want my data collected?
Weird question. The trouble with web automation tools such as Selenium and beautiful soup is that they require specific instructions to parse html. So it's difficult to auto scrape websites if several websites have a different html layout or any website simply decides to change it's html. Would it be possible to create a neural network model that could be trained to parse html as a human would? Or an AI for that matter that could break Google's captcha. I'm surprised that that still works given the advances in image recognition. If both are possible, the security on the web may quickly become a thing of the past.
What are the risks for generalized financial contagion in the traditional sector? I understand that crypto has been a way for wealthy individuals to hide funds from taxation, so if a bunch of wealthy dweebs lose their shirts I don't particularly care. On the other hand, I don't want to see this sparking another housing crisis or currency/FX crisis.
The servers hosting the data haven't been seized. Within a week there will be a dozen new domains for this information hosted out of Mozambique and Luxemburg or something. What a stupid waste of taxpayers money.
Ultimately what this means is that data transfer speeds will be faster, yes? (That's a question.) If most internet usage is direct to satellite and then satellite to endpoint this would mean that tier one hub networks would no longer be used. So laser replaces fiber for most of the distance, with a peer to peer satellite network that would make outages less common. Which would be both faster and more reliable. The only issue I see is for there to be an intentional or unintentional satellite debris event that would be self propagating, but that issue has been around for a while anyway.
To anyone interested I wrote a small paper on this and I'm looking around for commentary. It's mostly a philosophy paper, but it considers AI as well. I then go on to discuss moral and ethical questions.
Here's a link to the latest draft. I don't know that it will ever be good enough to be an academic paper, but maybe someone will find it entertaining.
Always looking for any substantive commentary. I'd be incredibly interested to know if anyone thought there might be a way to model A^ using a neural network that satisfied the conditions in the paper.
Thanks, I'll definitely look into those resources. I don't know if I'll go the route of sending in the paper myself as I believe a thesis advisor might be necessary to establish whether I'm on the right track or not without spending months on this. It's in a weird area between philosophy and AI and I might need two or more people to weigh in (not in the sending to journals stage, but in the research stage). I also don't know if building a demonstrable model is necessary for A^. More questions than answers.
I hadn't know about these journals though so that's a definite plus. I'll keep them in mind when looking where to submit. I should also mention my aim isn't to publish per se, but to know whether I'm right. I don't know if publishing the results exclusively (if they are significant) would be ethical.
Yes this is a very rough preliminary draft. The lack of references is a problem, although I cite the literature in philosophy for a number or people without pulling direct quotes [Timothy Williamson on modality for example]. The paper would need to be run through LaTeX, and the charts would need to be done in LaTeX. As to your suggestion I fixed some of the formatting, but it needs a complete rework later.
As to the literature for machine learning, I currently do not understand the state of the art. My understanding of the models is that they are systems of neural networks with varying degrees of complexity, but in essence are still neural networks (without A^ as described). I would need to read more on the literature or talk to someone who has as it's a complicated field. One issue is that I don't know that there is any one right model or system of neural networks that has been shown to work - some apparently do and some don't and it's been more of an experimental approach. So who to cite? Navigating the literature may be difficult.
With current machine learning (neural networks as I understand them) I believe that there is no "intelligence". But I think it's possible to create an A^ pattern and my paper provides bounding around this area to make it possible. Whether it is computationally efficient to train an AI from a bottom up approach (np-incompleteness) or if it would return results that would be communicable between the AI and a human if it did I do not know. I need to make A^ more formal and find a specific example, but I think it may be possible to create a model using an image classifier (not a neural network as is currently done with say hand writing OCR).
Yeah understandable. The paper is ambitious in that it starts from first principles in some areas by necessity, but I would have to have someone check for consistency and references to Husserl, Heideger, Lacan, Plato, Korzybski, Zizek, modality theory, vagueness, and more and that's just for philosophy. The neural networks is an issue in that I don't know if my references to A^ are addressed within neural networks currently, but I provide some evidence that they don't appear to be.
Mostly I just want someone to read my paper because the ideas are rather startling to me, but I don't know if this has been done before. If it's original work I think there are important reasons to have it published so everyone can read it at once, which is explained in the paper. But I may be completely off base.
To the extent that the lack of citation invalidates the conclusions of the paper? I don't know. I think that Korzybski's attitude towards maps, and several other areas where academics have been building on incorrect ideas (such as the theory of interest or that this would necessitate a Marxian dialectic in opposition), mean that original research need be done without necessarily building off of recent research.
Again, super ambitious paper. I could be entirely wrong.
My impression with arxiv is that it's for after the paper has been reviewed to make the information as widely as accessible as possible. I'm more interested in places that would review the paper in the first place.
This is the paper by the way. I haven't had anyone who has been willing to review it, I'm completely out of money and I just want to know if any of my conclusions are correct at this point. I don't know if this is the appropriate forum for review by academics per se but this is one of the more serious minded forums I've come across.