"AMD Silicon" covers wide range too:

1. circuit designers
2. silicon fab
3. supply chain

Our applications must be easily portable to this mode of operation

Need to move to open silicon, with trusted fabs

If cloud@home then the workload is rapidly relocatable, and so the customer may choose to run some base load on prem, have other load on cloud@home, and sell their spare capacity on cloud@home

cloud@home makes use of idle resources, and shares yours out too. you can have enough resiliency for standalone operation in a disaster, but can rent this capacity out when not in use

Show diagram of comparitive trust chains, latency, cost, fault tolerance of cloud configurations

https://www.theregister.com/2020/07/14/google_amd_secure_vm/ So your chain of trust can now be as short as "AMD silicon", Which paves the way for cloud at home type scenarios where renting off a home supplier has the same security guarantees as renting off a cloud supplier

event driven workloads designed for distributed storage - so the workloads would have to be designed specifically for cloud@home unless its just small little apps

"Mirror mirror, on the wall, who is the fairest of them all?"

Contents

1. Example project funding of Interblock
2. Example project funding of DOS

Software companies make up trillions of dollars on the stockmarket, but the best quality software is actually free - how can we connect these two extremes to something more sustainable, and how can we get more software, of a higher quality, faster, and cost lest money ?

"Mixing the pure with the impure, without making everything impure"

The work I have been doing lately could largely be described as side effect management. This is the means by which to introduce external data into the blockchain based system, in a way that does not corrupt the key blockchain property of reproducibility. That is to say, being able to rerun blocks thru different computers and verify that the same answer came out. Side effects are any external event - something that cannot be reproduced - a simple example is receiving an email. The converse is required too - being able to use the pure, clean, secure blockchain based calculations to trigger unreproducible external events. A simple example of this is sending an email.

So the work is about making something unrepeatable become something repeatable in a clean way, and to a lesser extent making something repeatable trigger unrepeatable events.

However during the work I was doing to send internet packets between Amazon and the browser the amount of code I was writing that was not blockchain related started to increase. The problem with this is that it is not reusing the code that is used for the blockchains, and so requires extra work to build and maintain, while losing the very helpful traceability properties that blockchain based calculations afford. Traceability becomes especially useful when the software is executing on remote servers.

This was the trigger for the work, but it is work that would have been required later anyway.

I have managed to get the underlying system of side effects working in prototype form, and starting from this morning I will be applying that to the internet transport ability. I am pleased with how the system is performing, and it meets my goals of being clean and simple and easy to fault find. An example of why the fault finding aspect is useful is that if you call me and say that at roughly 10am something weird happened in the app and you can't seem to make it happen again, we would be able to see the actual interactions the application made and very quickly be able to reproduce the problem, making fixing it much faster, and giving you reassurance that we did actually know what is going on, rather than having to ask you to keep an eye out for it happening again because we can't notice anything amiss.

When was it first invented

But the only truly zero liability company is a fully anonymous one.

Building customer trust

The only way to protect the privacy of your customers is to ensure they remain anoymous.

Switching cost for customers is high, so they will tolerate erosion of value before switching. Also company 1 and the livelihoods of all their staff are destroyed by a mass switch, when perhaps the majority of the company did good work. Then the process repeats again for company 2.

We propose a means by which there is really only one company, and the members cooperate at whim, and the consumers pay down whichever paths suit them best.

Outsourcing your identity management requirements.​

In such an ecosystem, the dangerous job of correlating identities need only be managed by a handful of services.

Exceeding quality of registered companies​

Companies do bad things all the time. Clearly being registered is no guarantee of good behaviour. With zero liability, the responsibily shifts onto the community to maintain opinion... This correlates with a shift in social mob justice of suspected bad actors.

Complete anonymity and business models that still work (hopefully!) even tho buy side and sell side do not know who each other are.

Making a functioning revenue stream when there is no force available - how to make money from open source software.

Transparency of companies is somewhat like how on a unix system, the operating system is largely readable by all users, but only writable by a trusted few. It helps the users to know whats going on.

We only care about the work generated. Work is hard, and full of thorns as the level of detail increases. Being zero liability and anonymous frees people to focus solely on the work, and frees consumers to look only at the value of that work to them, independent of any other forces that can alter perception away from cold purity of value. It is then up to the producer to choose to declare what they have done - this is optional, whereas under the current systems it is required, and criminally punished.

Moreover, we have implemented ways in which people can partially declare properties about themselves and their work rather than a full destruction of privacy. For example you can declare yourself a tax resident of x country for the projects you contributed there, and you can pay tax to that country, but you need not assosciate your name with it.

The bringers of novel knowledge are often unwelcomed, some violently so (references). For this reason, the Dreamcatcher network is designed such that any contributors of work or funds are free to stay anonymous, and at their wish only reveal as much as they wish. They can also release graduated information rather than all or nothing. Indeed consumers within the network also can operate this way, and share any portion of their consumption data on the network

Operating Activities​

1. Real life promotion (conferences, videos, friends)
2. Online Promotion (forums, emails, blogs)
3. Code commits
4. Issue posting
5. TESTING Amazon deployments using serverless.com tooling
6. PRODUCTION Amazon deployments using serverless.com tooling

the concept of the zero liability company only the code matters

Why the panic to create an invention machine ? One of the founders of Dreamcatcher Inc describes the burden carried by all those who know how to let the genie out of the bottle, and the consequences of delayed release.

5/22/19, 14:13 - 🦄: What about all the people dying around the world because of cures not yet invented? Is it our fault, if it makes a cure, and say 10 people died in the previous 3 months, and we're like "remember that three months we fucked around for?"

5/22/19, 14:15 - ☢️: Shit that’s a good point.

5/22/19, 14:15 - ☢️: Zero tolerance of delay.

5/22/19, 14:20 - 🦄: Fuck man. Is that an unreasonable burden to put on us?

5/22/19, 14:20 - ☢️: Erm, no...

5/22/19, 14:21 - ☢️: Unless you fancy quitting 🤣

5/22/19, 14:21 - 🦄: Just cuz we know what it is, now it's our burden to make it?

5/22/19, 14:21 - ☢️: Yes, precisely

5/22/19, 14:23 - ☢️: So cast your mind forward until you’re 80 years old. Starting to lose it but not yet. Pissing yourself occasionally. Multiple things fucking up your body. Guys born 30 years, 40 years after you now calling the shots. Then ask what you want to do with the time between then and now. How to get those guys harnessing all of those kick-ass concepts that are lost to the wind...

5/22/19, 14:24 - ☢️: Or leave it to them and hope someone else comes up with an Invention Machine...

5/22/19, 14:24 - ☢️: Never know, maybe someone will

5/22/19, 14:24 - ☢️: My guess is they won’t

5/22/19, 14:25 - ☢️: Because they’ll still be stuck by the stuff we’re trying to unstick

5/22/19, 14:26 - ☢️: So what, we have maybe 40 years useful left. 8800 days of work. About 70,000 hours

5/22/19, 14:27 - ☢️: And there goes another 4 from the time we called *** this morning

5/22/19, 14:27 - ☢️: Tick tock motherfucker…

5/22/19, 14:28 - ☢️: Then game over

5/22/19, 14:28 - ☢️: So, given all that. How long should we fuck around with the likes of ***?

5/22/19, 14:28 - ☢️: A year so far. Or about 2000 of those hours...

5/22/19, 14:29 - ☢️: It’s taken about 35,000 hours just to get to this point...

5/22/19, 14:29 - ☢️: One third gone

5/22/19, 14:30 - ☢️: Two thirds left

5/22/19, 14:32 - ☢️: “Despite this success, more than 3 million people die from vaccine-preventable diseases each year. Approximately 1.5 million of these deaths are in children less than 5 years old”

5/22/19, 14:33 - ☢️: So say there’s an invention the Invention Machine will find that stops 0.1% of that.

5/22/19, 14:34 - ☢️: Three months delay isn’t 10 people. It’s 750...

5/22/19, 14:35 - ☢️: And fucking around with *** is 3000...

5/22/19, 14:36 - ☢️: There’s huge value in getting something done by painting a vision. We can’t do this all the way just the two of us, but we can define the story others rally around

5/22/19, 14:38 - ☢️: “Just cuz we know what it is, now it's our burden to make it?” Yes, it is. Bring it on!

This is what 3000 people look like...

5/22/19, 14:41 - ☢️: Bear in mind you’re talking to the guy who thought he was about to kill 45,000 in 30 days at a 95% assurance of 70% probability in order to save 60,000,000...

5/22/19, 14:45 - ☢️: So, sense of urgency? Hell yes