FlashgodB

Caffeine ☕ early access users, post demo videos of fully onchain apps you've built on the Internet Computer / ICP by chatting with AI, and reply below with a link We'll distribute 7,500 ICP among the top entries, and give special gifts to runners ups Deadline: Sept 30 🍀

Blockchain must lead the AI 🧠 revolution & go beyond ERC20 tokens for fake decentralized AI: 1⃣ Host on-chain/network-resident (i.e. smart contract) AI models that are tamperproof, unstoppable & can be autonomous 2⃣ ❔⚡️ #ICP technical brilliance delivered 1 and 2 is coming.

Internet Computer is the only blockchain where smart contracts can: – be AI models – process big data – create web experiences – pay for their own compute Can people kindly suggest to @CoinMarketCap that #ICP should be on its "AI and Big Data" tab (we have asked...)

Crazy fact – 99.9% of people don’t understand what blockchains truly are. But, when you understand, you can see a future that's insanely bright 🌞 that runs far beyond defi and meme coins. I'm going to explain in this tweet... Three sections follow:- 1) The true nature of blockchain 2) The provision of near magical powers 3) The meaning of "blockchain singularity" Stay with me if you want to undertand! // The True Nature of Blockchain // To begin, you must grok that a blockchain is what can better be technically described as a "stateful decentralized network." If you're balking at getting your head around what sounds like hardcore technical term, don't panic – it's easy to understand, and the insight is well worth the amount of effort required. To advance on a short journey to understanding, we will start by considering, for a brief moment, the simple Morse Code protocol, and then add some modifications. As you may already know, Morse Code is a simple protocol for transmitting text over a link using a series of short and long pulses – audible beeps, flashes from a light, radio pulses, whatever... In Morse Code, 'S' is audibly communicated as beep, beep, beep, and 'O' as beeeep, beeeep, beeep. This means we can transmit "SOS" over a link by going beep, beep, beep <pause> beeeep, beeeep, beeeep <pause> beep, beep, beep. It's just a simple protocol that enables one party to send a stream of text – "data" – to another party. The protocol was developed to send text over electrical telegraph systems sometime after 1820. There would be operators who understood the protocol, who could both send text, and also transcribe received text, which their equipment transformed into audible noises for them (i.e. the beeps they listened to). It's a communication protocol that is "stateless." What that means is that messages are sent, and then the receiver does with them what they wish. Of course, in a sense, the operators on the ends of the line (the "peers") are maintaining some kind of state in their heads. For example, the operator hearing that the Titanic was sinking, might process that information in their head, and then decide to send back a message asking how long they've got before the ship will go down – but the protocol itself does not define how to maintain state, which would involve specifying how received communications should be processed and maintained. A similar situation exists with TCP/IP, the protocols that power the internet. When we connect to a remote person over Zoom, TCP/IP streams the video data over the network, and the Zoom app displays it. The people on the Zoom keep a state in their minds related to their interaction, which guides their conversation, but the internet is just streaming video data between them, and says nothing about what to do with the information contained. Early decentralized networks running over the internet, such as the BitTorrent file sharing network, used protocols that began to introduce the concept of some kind of shared state – albeit, one that is distributed across the computers of individual network participants, which is stored at their discretion. For example, the BitTorrent protocol makes it possible to get a listing of files stored on a peer's computer that they are sharing, and provides way for this knowledge to be propagated over its network. Crucially, the files ("data") shared by peer computers ("nodes") do not impact how BitTorrent's network protocol works, any more than the video streamed by Zoom affects what Zoom's protocol will do in the future. It provides a simple means for one peer to discover and request the files another peer is sharing, and what it knows about files shared by other peers, allowing for propagation of this knowledge. It enables file sharing but is oblivious to the files being shared. The files are not part of a state the network protocol itself maintains. Bitcoin rang in the changes as the first "stateful" decentralized network – something whose revolutionary nature well exceeds the concept of cryptocurrency. Going back to two people communicating using Morse Code: imagine if that protocol was updated, so that each party had to keep a record of the text received from the other, and that new Morse Code protocol commands allowed each party to request the other send back sections of text they previously transmitted – and that correctly behaving participants had to fastidiously record received text in case they were asked for it. Now we are moving towards a weakly "stateful" communications protocol involving just two peers. It's not decentralized since there are only two parties, and there's no way to know if text returned on request is correct (unless you kept a copy..). Now, imagine if the protocol was made decentralized, and when a new party joined, they could request all the text that had been previously transmitted from all existing peers. Assuming peers regularly updated each other about the transmitted texts they had seen, this would allow a rough *shared* corpus of text to be stored by the network. Whenever new text is transmitted, this would be shared amongst the peers in the network, and thus, future requests to return the corpus would include it. If the protocol requires that peers share texts they have seen, and the peers behave correctly, then we can say that the shared corpus (a "state") is being maintained by the network. But there are two problems: 1) the corpus maintained by the peers is inconsistent, since they transmit messages, and see shared transmitted messaages, at different times, and 2) peers can lie, sharing texts that were never sent, causing the network's state to become corrupted. There is no "source of truth" that describes what the correct state maintained by network is. We can make a few mild improvements without introducing a trusted intermediary – such as a server that might act dishonestly or go down – to maintain the corpus. For example, when a peer is asked for the network state, the protocol rules could require they sort the text messages they have accumulated by the identity of the sender, the time they were sent, which will make the state returned by one peer look more similar to the state returned by another peer, but still will not make them consistent, because the peers see messages at different times. And there's no obvious way to deal with adversarial ("faulty") nodes that share messages that were never transmitted, which will corrupt the state. This is where the fields of Byzantine Fault Tolerant protocols comes into play. These involves complex math and algorithms, but the good news is that you only need to know that math proofs have been created to validate they work – you don't need to understand their works. These protocols can guarantee that if a sufficient portion of peers follow the protocol rules (are "correct"), then a completely consistent network state can be maintained, and moreover, that faulty peers that don't follow the protocol's rules cannot corrupt this state (here we assume that transmitted messages are now signed using cryptography to prevent forgeries, and are really talking about the construction of a *consistent* state that everyone can agree upon). Such protocols have to address the network's participants seeing messages at the same time, which is unavoidable. They address this by creating a shared state that progresses through an agreed sequence of steps (or "state transitions"). Going back to Bitcoin, this is of course what we see in its sequence of blocks, which defines progressive updates to its UTXO set it uses to describe the various balances of bitcoin at different wallet addresses, and how it can be accessed. This shared network state is the Bitcoin "ledger." In this network, if Alice transfers some bitcoin to Bob by sending her "transaction" to her peers on the network, anyone in possession of the block containing her transaction can see that the transfer occurred, because the sequence of states is consistent, and that Alice cannot now send the same bitcoin to Carol and that Bob is thereafter in possession of the bitcoin. Bitcoin was the first true "stateful decentralized network," capable of creating a shared consistent state without a trusted intermediary, which resists adversaries – through the application of protocol math. Beyond the invention of cryptocurrency, Satoshi had created a decentralized network that hosted a ledger, where the data and all related computations lived *within* the network's protocol interactions. In this new type of network, peers might cache copies of the shared state, but the protocol defines exactly what the state should be, and maintains constraints regarding how it can be updated. Because the state, and computations that update the state, live within the protocol, and are restricted by the protocol, they do not depend on special trusted peers (computers) keeping definitive records. To corrupt computation and state, you must corrupt the mathematical protocol. This is completely different to BitTorrent, say, were a protocol command issued by one peer to another simply obtains state that it maintains. Byzantine Fault Tolerance means that If a peer is adversarial/faulty (or "Byzantine" in the arcane speak of distributed computing), then their incorrect responses are detectable and can be discarded by correct peers, and whatever these bad peers do, they cannot corrupt the state shared among correct peers. So long as the math is good (which requires proofs, that many protocols fail to produce), the network is invioable since math is not subjective and 2+2 always equals 4, not 5. Stateful decentralized networks also require other features that I will not cover in depth here. For example, participation needs to be gated some way to prevent an adversary adding so many peers that they break the "fault bounds" of the protocol math involved. This is what is referred to as "Sybil resistance," and a public network can create it in different ways. For example, Bitcoin uses proof-of-work, where costly hashing stands in for voting on a peer-by-peer basis. Ethereum uses proof-of-stake, where staked/locked ether is a stand-in ("voting by capital"). The Internet Computer has a decentralized permissionless governance system which defines the peers, which consist of standardized hardware nodes. In this network, these nodes must produce the correct number of valid blocks required of them by randomness, in a system of proof-of-useful-work, to continue participating. Despite using different approaches taken to "Sybil resistance," all three networks are "stateful decentralized networks." To understand the utility they try to provide, we can divide them by blockchain generation. Bitcoin is a First Generation network, where the network state is a simple cryptocurrency ledger that can be updated. Ethereum is a Second Generation network, which adds smart contracts (i.e. software and related data) to its shared state. And the Internet Computer is a Third Generation network where an evolved form of smart contracts can do general-purpose computation that can be used to create something like a social network, entirely on the network, without need for traditional IT. See where we're going? // The Provision of Near Magical Powers // Bitcoin showed us that shared network state and computation could be 1) tamperproof, which means that state can only be updated by computation initiated by correctly signed transactions, without backdoors and security that is inviolable, 2) unstoppable, which means that so long as there are sufficient peers running, it has 100% resilience, and 3) autonomous, in the sense that state and related computations do not rely upon a person, organization, or other intermediary. However, token ledgers are just the start of the story. Ethereum showed us that the technology generalizes to computation beyond simple ledgers by hosting smart contracts. For example, because the Uniswap decentralized exchange is hosted by the network, it is also tamperproof, meaning no cybersecurity systems are required to defend its state and operation, and every computation always only involves its correct logic/code, and is legally initiated. Furthermore, so long as Ethereum continues to run, its code can always be invoked, and it is therefore "unstoppable." Uniswap also demonstrates the power of autonomy: each particular version installed on Ethereum is unmodifiable, such that no individual controls the code and therefore cannot surreptiously modify logic to steal user funds. Uniswap has these properties because it exists within the Ethereum network's protocol interactions, rather than on a server or cloud. To subvert it, you must subvert Ethereum's protocol, which is mathematically secure (at least assuming the current Ethereum protocol is supported by mathematical proofs @VitalikButerin?). Stateful decentralized networks, then, provide utility that includes them hosting data and computation that is tamperproof, unstoppable, and supports autonomy. We can also add other useful properties such as being borderless, which is important in an increasingly globalized world, and in the case of smart contracts, composability. I could go on waxxing about whese wonderful properties, but — Standing back, and setting aside long-held assumptions about speed, scalability and efficiency limitations, its already obvious that stateful decentralized networks are really a completely new form of compute platform, and that on closer inspection, his new form of compute platform provides potentially overwhelming advantages. When you build a service such as a defi system, social network, or enterprise system on a stateful decentralized network, the service becomes tamperproof, and you don't need to protect it using fallible cybersecurity systems like firewalls, or anti-malware, because its data and computation lives inside mathematical protocol interactions, which precisely restrict what can and cannot happen, and its unstoppable. // The Meaning of "Blockchain Singularity" // The global cost of cybercrime is predicted to exceed 10 Trillion dollars in 2025, which will be 35%+ of the GDP of the United States. Cybersecurity is becoming a major problem of our time, like war and climate change, and is just one of the key problems that stateful decentralized networks, in their many manifestations, can solve for. Yet few have zoomed out. This is partly because of widespread misconceptions that blockchain networks have fundamental technical limitations, partly by a widespread preoccupation with speculative instruments like NFTs and memecoins within the industry, and partly the difficulty of seeing that pervasive and prevailing challenges like cybersecurity, might be solved. However, as speed, scalability, efficiency and other limitations are rolled back, stateful decentralized networks will increasingly provide a superior form of alternative tech stack. Instead of building system and services using server computers, clouds, virtualization, orchestration software, database servers, application servers, web servers, firewalls, anti-malware, and all the rest – humanity will build systems and services that are secure by default, which never get corrupted, which never stop running and need restarting, and in the world of web3, we will build things like autonomous social networks that run under the control of DAOs that provide transparency and deliver 100% ownership and full sovereign control to internet communities. Stateful decentralized networks – blockchains – will function as unstoppable decentralized and secure "serverless clouds," where anything can be built in forms that are vastly more secure and resilient. These clouds will also finally provide a route out of the complexity of the current IT stack. This is also important, because the world currently spends around 2 Trillion dollars a year on IT personnel, who now spend 95% of their time just chasing complexity. Meanwhile, it's clear that an Ethereum smart contract, say, exists within a highly simplified seamless environment, and for example persists its data without the need for things like database servers. Blockchain will become a stack that drives incredible IT personnel productivity through simplification. Third Generation blockchain chases these immense opportunities. It's a new paradigm that sees blockchain as humanity's new IT stack, upon which *any* system or service can be built with immense advantages. In the future, Third Generation blockchain will provide unimaginable advances in IT, solving for key problems of our time such as security, resilience and IT productivity, which many incorrectly assume are intractable. As I have shown in recent video demonstrations, you can run AI models as smart contracts on the Internet Computer *today* – having them do tamperproof and unstoppable inference. Third Generation blockchain will therefore play a key role even in this new arena of tech – in the future, AI models run by organizations will have access to their most sensitive data, and running them on blockchain can solve for the risk they get hacked. We will also come to heavily depend on AI in numerous ways, and therefore it's imperative that AI models have massive resilience against infrastructure failure, which blockchain again solves for. Plus in the realm of web3, oftentimes it will be essential that AI runs autonomously, which only blockchain can enable (for example, imagine an AI that verifies the Solidity code of Ethereum smart contracts, and certifies them; it's essential users will know the certifications come from the AI, not from some person, organization or other intermediary that they then have trust). At DFINITY, a large and eminent team of cryptographers, computer science researchers and engineers has been pursuing the Third Generation blockchain paradigm for years, and continues to make rapid advances – even though the Internet Computer network you see today already represents more than 1000 person years of R&D. We plan to continue this effort forever. We understand that is blockchain is stateful decentralized networking, the magical powers this essence unlocks, and why that makes blockchain the future of compute. As the technology advances, Third Generation blockchain is reinventing compute in realtime. Web3 social networks, social media and games are already running on the Internet Computer end-to-end, without the help of current/traditional IT such as server computers, cloud services and traditional platform software like databases and web servers. These services already host many thousands of users, showing how Third Generation blockchain networks can scale, and many of the services keep tokens inside, yet run hack-free without any traditional cybersecurity protections at all – speaking to the security revolution underway – while running autonomously under the full control of community DAOs. Third Generation blockchain is decentralized cloud that hosts tamperproof, unstoppable and optionally autonomous systems and services, with additional functionality that allows hosted service to for example to serve web experiences directly to users. It's performant, scalable and efficient, and rapid improvements continue on all key axes. I'm proud to say that this revolution in compute began within the blockchain/crypto industry, and is one of many gifts it has given the world. However, Third Generation technology will now quickly spread to the enterprise and government IT sectors, where system security and resilience, and the productivity of IT personnel are critical concerns. The UTOPIA (Unstoppable Tamperproof Open Platform for Independent Autonomy) venture will make ICP technology available for enterprises and governments that cannot build on a public network, because they need to control which peers/nodes process their data, allowing them to create the Internet Computer equivalent of intranets, which nonetheless interoperate with the public internet. Our purpose at DFINITY, and other Third Generation networks when they start to appear, is to reinvent compute itself. Our purpose is to get the vast majority of humanity's systems and services running on blockchain. The scope and importance of this endeavor is similar to that of AI, which field will also benefit enormously from the reinvention of compute. Sometimes our industry has become myopically focused on the speculative side of blockchain, and blockchain narratives designed to lure in speculators. We want to open people's eyes. Our industry is not just about that. Our industry is about the almost magical powers that stateful decentralized networks unlock within the world of technology. In the past, I have talked about "blockchain singularity." The term does not infer a future where almost everyone in the world will come to own a memecoin, NFT, or participate in restaking – or anything like that. It means is that the vast majority of systems and services that humanity uses will one day run on blockchain – in a process that will occur progressively over years, as our networks inevitably eat into the multi-trillion dollar tech pie. This is what I'm talking about when I say "smart contracts will eat the world" in presentations (riffing off a16z's expression that "software is eating the world"). Third Generation blockchain is our future tech stack. Once the nature of stateful decentralized networks is known, this much becomes obvious. The Third Generation blockchain will deliver exponential impact within the world of IT, through seminal advancements such as solving the cybersecurity challenge. So that's it for this tweet! Thank you for reading if you stayed with me this far. Join us reinventing compute using Third Generation blockchain that is stateful decentralized networking without limits. Everything on blockchain ⚡️💪👊

A world first: a demonstration of AI running *on* blockchain as a smart contract 🧠⚡️ The Internet Computer is used – the world's first 3rd gen. blockchain #ICP. AI will become the beating heart of our web3, multi-chain world, and this is only my first demo. Code will follow shortly. This is running on DFINITY's Internet Computer testnet, but you'll be able to take the code and run it on the public network as the NNS is expected to up the per-transaction instruction limit in the coming days. Some important notes. The inference engine used has not been optimized, and we will show vastly greater efficiency in subsequent demos, where the AI runs even faster, and consumes less gas/cycles. We will also propose to the NNS that smart contracts have access to SIMD instructions – which we have determined are deterministic – unlocking vastly more speed and efficiency. Lastly, currently the actor smart contracts ICP hosts run inside a 32 bit environment, which limits their main memory to 4GB. Within the next couple of months, we expect the Internet Computer to move to a 64 bit environment, allowing smart contract memory limits to be raised much higher – allowing for models with far more parameters, and thus power, to be run. Those of you who have followed my posts know that hardware optimization support is planned. Firstly this will involve WASM smart contracts shipping out matrices for processing by the CPUs on existing node machines (another advantage of the Internet Computer running on sovereign node hardware). Secondly, we plan to propose a new public node machine specification, which node providers can build to, for use in subnets specialized for hosting AI smart contracts, in which each machine will incorporate several GPUs. Naturally, this will be packaged with other technological developments that ensure smart contract determinism. This is part of our mission to enable powerful LLMs to run as smart contracts on the Internet Computer – in a forms that are tamperproof, unstoppable, and optionally autonomous (including under the exclusive control of DAOs). Our vision: you will be able to have a chat with a smart contract. A smart contract will be able to coordinate your organization (see my earlier tweet about a "Delphi"). ICP smart contracts will be able to audit Ethereum smart contracts, kitemarking those without backdoors and reentrancy bugs. And smart contracts will be able to do KYC autonmously, matching faces to driving licenses, and more. We will deliver new SDK enhancements, and work with partners, to turbocharge crypto AI developers. Secure and unstoppable AI and third generation blockchain will be two sides of the same coin. Security and AI will be indivisible. AI models will be traded as NFTs. Thanks to Internet Computer's chain key (trustless multi-chain) capabilities, all blockchains will be able to leverage AI smart contracts. The future is beyond exciting. Thanks for watching. Can't wait to give you more demos!

Internet Computer/#ICP is leading these new fields 🔥 Full Stack Decentralization/#FSD - web3 services that are 100% on-chain Decentralized AI - LLMs running as smart contracts (chat with them, trade them as NFTs...) DePIN - Sovereign internet infrastructure, not cloud

The American establishment is wary of the #NetworkState and citizens belonging to web3 internet communities. But, on a global scale, might this lessen the dangers of revanchist dictatorships that could cause WWIII, and be a way democracy, liberty, globalization and e/acc wins?

What bothers me about a lot of L2 scaling solutions, including zk solutions, is that they are trying to solve the problems of yesterday's thinking, using more of yesterday's thinking. Just design a scalable blockchain, damnit #ICP

#ICP leads the "DePIN" category for networks backed by dedicated hardware 🔥💪 Weekend thought: propose a new "FSD" category for networks powering "full stack decentralization" i.e. 100% on-chain services Internet Computer would also lead, but as only member

Today, #ICP is powering "open internet services" that run 100% on-chain: – Controlled by community DAOs, – tamperproof and unstoppable, – accounts acting as wallets, – with trustless multi-chain Will mainstream understanding of Web 3.0 include this new paradigm within 18 months?

#ICP security involves different thinking vs proof-of-stake networks, so direct comparisons are hard. Differences include: 1 "Deterministic decentralization" ⬆️ vs simple Nakamoto Coefficients 2 Games of life ⬆️ vs small-game fallacies 3 Hardware ⬆️ vs liquid stake Big thanks to Justin Bons for his willingness to debate the ICP community on this subject – these days theoretical debate is sorely lacking – but he does not cover these differences in his analysis. He therefore provides a good opportunity for me to draw attention to the differences, and explain them to those interested: /* Deterministic decentralization */ Decentralization, generally speaking, is a nebulous term that is difficult to measure – should we measure the decentralization of the Bitcoin network by the rather small number of mining pools that might double-spend by colluding, or by the much larger minimum number of individual miners that would need to collude, upon the assumption that miners would leave any pool involved in an attack? Or, should we measure the decentralization of Ethereum using the number of individual virtual validators nodes running in the cloud, or by the number of Big Tech clouds that control them? These are tough questions. The Internet Computer uses a decentralization framework, which underpins the security of its protocols, known as "deterministic decentralization." This aims to calculate and create decentralization both with more certainty (more security) and also with less replication (more efficiency). It is more nuanced than basic "Nakamoto Coefficient" thinking. The system is made possible because the network runs under the direct and exclusive control of an advanced permissionless DAO that is integrated into its protocols called the NNS, or "Network Nervous System," which currently has more than $3b of ICP staked and huge numbers of participants (it's the biggest DAO community around today). The NNS adopts/rejects proposals that do many things, two of which are particularly relevant to decentralization: 1) it hands out "node provider" profiles to those who ask for them, and 2) scales the network by forming and configuring subnet blockchains by adding and removing node machine hardware operated by different node providers. Node providers obtain profiles by submitting proposals that specify the data centers they will be installing node machines within, which information then gets associated with their profiles. Another type of proposal, which anyone can submit, forms new subnets from existing nodes, or adjust subnets by adding/removing nodes, with the requirement that "deterministic decentralization" is observed. This involves combining node machines from 1) different node providers, 2) installed in different data centers, which are 3) in different geographies, and 4) different jurisdictions. The purpose is to create subnets which would require a higher degree of node provider collusion to break, and which are also resilient to e.g. a nuclear bomb hitting data centers in a geography, or a government/regulatory area attack executed within a jurisdiction such as the EU. Some subnets also have special purposes, such as those hosting e.g. bitcoin, and have higher replication factors as needed – security is on a cost/benefit curve and tailored precisely to need. It is important to compare this to traditional proof-of-stake practice, where validators are anonymous, and mostly run on Big Tech's cloud services. For example, the Hetzner cloud suddenly shutdown 40% of the Solana network, more than 1,000 of its validators, after deciding it didn't want them. By implication, it could also also have taken full control of the network. Further, consider that vast numbers of cloud validators on Ethereum are in practice sometimes created by one operator behind the scenes. Arguably, the cloud providers are the nodes, not the validators they host. Deterministic decentralization does not depend on having huge numbers of anonymous validators running in the cloud, and I would argue provides more dependable decentralization that is also certainly much more efficient. If required, the Internet Computer's network protocols allow for very large numbers of nodes to be combined into subnets. Where they are not, it is because analysis performed wrt deterministic decentralization indicates that it would be unnecessarily wasteful. /* Games of life */ I don't always agree with Nick Szabo's thinking, but one of his pieces that I think is INCREDIBLY insightful and generally important for the crypto industry, and often overlooked, is his post titled "Small-game fallacies." unenumerated.blogspot.com/20… Small-game fallacies often occur in proof-of-stake modeling, which assumes that validator behavior is determined exclusively by incentives existing within the narrowly defined constraints of crypto economic systems e.g. that anonymously deposited stake ("collateral") is the primary determinant of validator behavior, including cheating/not cheating. This is incorrect. Szabo draws attention to the fact that human behavior occurs within the much larger "games of life." For example, most validators won't cheat, not because of fear of losing their stake, but for reasons of personal integrity, professional pride, and fear of retribution (both from others in their industry seeking revenge, and from legal systems that with to punish them for their crimes). These incentives can exert much greater forces than anonymously staked cryptocurrency. The Internet Computer considers the larger games of life by ensuring, among other things, that node providers (which are typically companies) identify themselves. Those colluding to defraud the network will be subject to much more than the loss of their stake. Indeed, they can be subject to the full force of the law around the world. As the assets hosted by networks become increasingly valuable, it becomes harder and harder for proof-of-stake advocates to claim that they are sufficiently protected by validator stake. The larger games of life are required. /* Hardware stake */ The Internet Computer runs on sovereign hardware called "node machines," which are built/purchased by independent "node providers" who install/run them from independent data centers around the world. In the "proof-of" terminology, it is "proof-of-useful-work" (PoUW). This hardware-based approach is diametrically opposed to proof-of-stake, which typically sees validators running on Big Tech's cloud services, run by anonymous operators, who have joined them to the network staking some amount of cryptocurrency. As per the preceding section, the thousands of validators running on Amazon Web Services should really be seen as one node, for reasons highlighted by the Hetzner incident, but that's beside the point – what we need to compare here is the *types* of stake involved in proof-of-stake vs Internet Computer nodes. An ICP node is not joined to the network by staking ICP tokens, but rather by merit of being standardized hardware that can keep up with the nodes it is combined with (and thus not get expelled). This standardized hardware is expensive. It costs $20k+ per node machine (and those wishing to support optimized AI smart contracts will likely find the specs for those nodes will probably exceed $100k). This hardware is of course a form of stake that the node providers put at risk, and it is not one that is not liquid. The cost fully includes both the cost of acquiring and installing the hardware, and the hosting contracts with data centers. If a node provider is kicked off the network for cheating, or running substandard hardware that can't keep up with the block rate, in practice they will have problems reselling their accumulated hardware, will be unable to recoup setup costs, and will often be stuck with contracts for co-location and bandwidth. By contrast, a proof-of-stake validator node on Amazon can be spun up and down on a button click. Nothing is staked apart from the cryptocurrency. However, that can be hypothecated using liquid staking – in short, what is actually staked is a much more wooly subject. While, as explained in the previous section, we believe that the larger games of life trump the small game fallacies of narrowly defined proof-of-stake crypto economic systems – not just in theory but technical practice – we also believe that it's often hard to determine what is actually staked when the stake is cryptocurrency. We would argue, as with many things in crypto theory, that the truth is far more nuanced that the big brassy theory boilerplates justifying the design of major proof-of-stake networks would suggest. In summary – the stability of hardware stake, and the forces created by the larger games of life unleashed via intelligent decentralized governance that carefully combines hardware operated by identifiable node providers in a scheme of deterministic decentralization, allows decentralization to be measured and tuned more accurately with far greater predictability, and therefore safety.

ICP is designed to revolutionise compute, and decentralization is key to its design. The network runs under the direct control of an advanced DAO called the Network Nervous System (NNS). Anyone can vote on proposals submitted to the NNS by locking up ICP tokens to create "voting neurons" (they are called neurons because they can be configured to vote automatically by following other neurons on a topic-by-topic basis). Large numbers of proposals are being processed all the time, and they can be submitted by anyone. Some are related to general governance, and simply coalesce community opinion, for example regarding questions of network design. Other proposals actually cause the network to do something in a fully automated way. For example, they can upgrade the node software running on node machines around the world, or adjust one of its internal subnet blockchains – the decentralized NNS directly manages and evolves the network. The purpose of ICP is to revolutionise compute. You can build a social network on ICP, 100% on-chain (see OpenChat, oc.app, for a great example of what's possible). This raises the challenge of decentralization Vs efficiency. The automation of the network by autonomous decentralized governance is essential to the network's application of "deterministic decentralization." Deterministic decentralization is how the network solves the efficiency challenge: the Internet Computer is created by special "node machines" that are currently run by about 120 different "node providers" around the world, which nodes the NNS organizes into subnet blockchains in ways that create high-certaInty large Nakamoto coefficients, while minimizing replication, and thus cost. When creating or reconfiguring a subnet blockchain (e.g. by adding or removing node machines), in applying deterministic decentralization, the NNS takes account of a lot of information. It aims to ensure that the node machines that make up a subnet are a) owned/operated by independent node providers, b) installed in different data centers/co-location facilities, which are c) in dispersed geographies and d) different jurisdictions. This can provide much more robust decentralization than anonymous PoS validators running on Big Tech's clouds, while – crucially – massively reducing replication cost, which is essential to the Internet Computer's mission of revolutionising backend compute on blockchain. Last note: the configuration of subnet blockchains by the NNS also allows for the creation of specialist subnets e.g. recently a subnet was formed entirely from node machines running in the EU, which will allow canister smart contracts uploaded to the network to be tagged "GDPR-compliant." Last last note: ICP subnet blockchains are true subnets, not standalone blockchains. They are combined into a single seamless blockchain environment, and are invisible to hosted smart contracts and users. Last last last note: technically, the reconfiguration of subnets by the ICP protocol (Internet Computer Protocol) on direction of the NNS is highly complex, and involves a lot of specialized cryptographic protocols, including a "non interactive distributed key generation" protocol, "key resharing" protocols, and many other things. TLDR; the Internet Computer is very decentralized by virtue of highly advanced unique crypto technologies

Just green lighted work to create an #ICP ERC20 twin on Ethereum using simple impl. Expect will be available *before* protocol-level "chain key crypto" Internet Computer <> Ethereum network integration (massive work on that now)

Expecting Internet Computer #ICP to enable LLM AIs to run inside canister smart contracts 🧠 thanks to new node types. Well be amazing for e.g. creating a medical AI as contracts will make provable: 1) the model code, 2) the training data, and 3) that a given prompt produces the provided response/so no human meddling. Compute will be more expensive, but won't matter for such applications. Likely 4X or 7X replication. Crypto buffs will note that modern AI relies on floating point GPU hardware, which is non-deterministic, and would cause a chain to fork. But the good news is it can be transformed to rely on deterministic integer arithmetic, which is also supported by new GPU hardware. Like a lot of work DFINITY pursues, this is non-trivial, but hoping we can at least get a strong PoC onto the network in 18-24 months. No guarantees. Meanwhile, we are aiming to help get the network supporting much simpler "AI compute units" this year. These won't run as smart contracts that can support creation of fully trustless AIs. However, they will run on a decentralized public network, use enclave tech, maximize efficiency, be natively integrated with Web3, have significant computational power, and provide unique capabilities to be explained later, which will prove very, very important... Blockchain will host AI and imbue it with special properties. #BuildOnTheNetwork