It's not a fair comparison. Money laundering controls likely have also been implemented since the 19th century in some way shape or form. Therefore, that 2-5% figure also includes the effect of those controls. This might just the natural equilibrium below which the juice is no longer worth the squeeze.
Access to information and interconnectedness have also increased exponentially so it might make sense that the costs have increased proportionately.
Finally 200B/year would not solve world hunger if nefarious actors grift away most of that money (especially because you wouldn't be able to track them with out some form of KYC)
You also have to ask if money laundering is bad enough of a crime to warrant all of the overkill responses to it. At some point the medicine becomes worse than the disease. Lawmakers tend to get into these inescapable loops:
Crime X exists -> Try solution Y -> Crime X still exists -> Try harsher solution Y0 -> Crime X still exists -> Try harsher solution Y1 -> Crime X still exists -> Try harsher solution Y2...
I think the problem here might be that while Crime by its definition might be the same, the way that it is executed changes drastically so Solution Y/Y0/Y1/Y2 are not required to exist in terms of severity (very likely solution Y was severe enough if you were tried and convicted of the crime) but they are needed because the way Crime X is executed changes (especially this Crime X)
The times is not really the same newspaper anymore.
Fun and eye opening thought experiments:
1)find an recent article you are an expert on and critically analyze it’s content from correctness and accuracy. (Gell man amnesia effect).
2) pick a common article (crime or accident) from recently, go back 30+ years and find another similar article. The compare the content for content, style, and tone.
Unless you're running your own local DoH server, it's just privacy theater, and not even good theater. It's the same bogus argument made in favor of retail VPNs. "Encrypt your traffic so your ISP can't spy on you!" Meanwhile failing to recognize that you're just moving the spying agency to some free service owned by--oh, let's look at the top providers... Cloudflare (not great), Brave (advertising company), Google (Bad)... Need I go on?
Perhaps you've forgotten that old adage: if you're not the customer, you're the product.
Security theatre is far too strong. It’s not perfect but it does solve real world, actively exploited privacy flaws, such as ISPs monitoring and selling insights from DNS, NXDOMAIN hijacking, etc.
Your internet traffic is all MITMd by Cloudflare anyway. Using their DNS barely moves the needle.
I run a local DNS server that has its own blocklists, internal DNS records, etc. I imagine most people who might bother to deploy an enterprise browser policy do something similar.
You’re describing the dump and pump strategy that china has historically used across a number of industries. Stands to reason that this is what is going on.
You don't have to build the entire software factory at once. You don't have to mastermind the whole future system, instead you're actually stacking and layering these small, isolated problems.
I think that's a really good approach to start getting value tomorrow or this week without saying, "I'm going to revolutionize how we ship."
It's just:
1. Find places where you can use agents.
2. Figure out where the right leverage points are for humans and where the right leverage points are for agents.
3. Just start building those things and plugging them into each other.
One day you'll wake up, and 80% of all of your stuff is automated.
Totally. Also to add to that. I think (like any factory) you need to build in systems to allow you to have observe and service the machines make process changes as needed.
You're the foreman/factory manager. You can optimize the factory over time upgrade machines, place machines closer or add conveyor belts for more productivity.
The only think that slightly worries me is that the Labs are clearly incorporating the best in class logic from their datasets.
I prototyped something like this for fun a long time ago. Treating s3 like a bucket of blocks seemed intuitive way build a scalable filesystem. Arguably ceph and luster are doing something similar except with a seperate metadata servers to serve the hotter content.
I think the critical thing you will need to explain is durability and loss window. Making some guarentees on failure modes would go a long way towards making me believe i can run operations on something like this.
With AI you should be able to do some exhaustive testing both for load, power loss, server loss, etc. Anxious to see the potential results
You shouldn’t expect frontier models to work on medical imaging. There is much more that goes into building a medical imaging product. First and foremost is data. Medical imaging datasets are not prevalent one the public internet at the scale necessary to have good performance on medical imaging tasks especially MRI. Also the labels are super noisy.
This is completely different than asking for general medical reasoning which is more derived from papers, public standards and textbooks.
Look into the Eigen library. They use template meta programming to chain linear algebra operations in a way that the compiler should be able to optimize memory layout and kernels for vector instructions. Might give you some ideas.
It is, I guess it's depending what you wish to achieve. I'm a independent software consultant based in Germany (no employees etc, also not selling any SaaS or anything) and I've got a liability insurance to cover any mess ups.
Ran the same query and there is a ton of stuff, but it looks like it's reasoning through the ambiguity of the sentence. It still gets the right answer. Moreover, if we consider the FLOPs expended to get to the answer, and compare that to opus, I think it's still a net win.
My hunch is that Opus scale models probably have shortcuts encoded into the model that handle these ambiguities cases, wheres this model has learned a program to reason through the edge case (crystalized vs fluid intelligence). Remembering that probablity (frontier) vs calculating it on the fly (vibethink)
Access to information and interconnectedness have also increased exponentially so it might make sense that the costs have increased proportionately.
Finally 200B/year would not solve world hunger if nefarious actors grift away most of that money (especially because you wouldn't be able to track them with out some form of KYC)
reply