{"action":"Motion","action_nns_function":null,"classification":"constructive","deadline_timestamp_seconds":1774326461,"decided_timestamp_seconds":1774326461,"derived_proposal_information":null,"dfinity_proposer":true,"executed_timestamp_seconds":1774326461,"failed_timestamp_seconds":0,"failure_reason":null,"id":121076,"known_neurons_ballots":[{"id":"27","name":"DFINITY Foundation","vote":0},{"id":"14231996777861930328","name":"ICDevs.org","vote":2},{"id":"428687636340283207","name":"CryptoIsGood","vote":2},{"id":"10843833286193887500","name":"Anvil","vote":2},{"id":"55674167450360693","name":"ICPL.app","vote":0},{"id":"12860062727199510685","name":"ysyms","vote":2},{"id":"8959053254051540391","name":"The Accumulators\u2019 Neuron","vote":0},{"id":"6362091663310642824","name":"RawTech Venture","vote":0},{"id":"8777656085298269769","name":"Paul Young","vote":0},{"id":"5728549712200490799","name":"ICPMANUAL","vote":0},{"id":"13538714184009896865","name":"8YearGangDAO","vote":2},{"id":"13765488517578645474","name":"Isaac Valadez","vote":0},{"id":"5944070935127277981","name":"Krzysztof","vote":2},{"id":"11053086394920719168","name":"Nicolas.ic","vote":2},{"id":"16335946240875077438","name":"Smaug\u2019s Neuron for Retail Investors","vote":0},{"id":"11595773061053702367","name":"ICLight.io","vote":0},{"id":"16737374299031693047","name":"Taggr Network","vote":1},{"id":"10323780370508631162","name":"Sonic AMM","vote":0},{"id":"6914974521667616512","name":"Rakeoff.io","vote":2},{"id":"17682165960669268263","name":"OpenChat","vote":1},{"id":"4714336137769716208","name":"ELNA AI","vote":0},{"id":"1767081890685465163","name":"GEEKFACTORY","vote":1},{"id":"2776371642396604393","name":"ICP Hub M\u00e9xico","vote":0},{"id":"5132308922522452058","name":"ICP Hub Poland","vote":2},{"id":"7902983898778678943","name":"Jerry Banfield","vote":2},{"id":"433047053926084807","name":"WaterNeuron","vote":1},{"id":"1100477100620240869","name":"ICP Hub Bulgaria","vote":0},{"id":"7446549063176501841","name":"Gold DAO","vote":1},{"id":"8571487073262291504","name":"NeuronPool","vote":2},{"id":"16459595263909468577","name":"LORIMER \u267e\ufe0f \ud83d\udc36","vote":2},{"id":"16122208542864232355","name":"B3Pay","vote":0},{"id":"3172308420039087400","name":"ZenithCode","vote":0},{"id":"12093733865587997066","name":"Aviate Labs","vote":0},{"id":"4713806069430754115","name":"D-QUORUM","vote":0},{"id":"16673157401414569992","name":"Yuku AI","vote":0},{"id":"5371276303191057244","name":"1e0","vote":1},{"id":"33138099823745946","name":"CO.DELTA \u25b3","vote":2},{"id":"12977943926061800402","name":"DragginCorp","vote":2},{"id":"1637856224910350276","name":"Gwojda","vote":2},{"id":"2334327054503903846","name":"Thyassa","vote":2},{"id":"10389857423811734339","name":"\ud83c\udf2e TACO DAO Neuron","vote":0},{"id":"18422777432977120264","name":"\u03a9mega-reject","vote":2},{"id":"18363645821499695760","name":"\u03a9mega-vote","vote":2},{"id":"2947465672511369","name":"\u03b1lpha-vote","vote":2},{"id":"4069412273344914316","name":"Cybersecurity","vote":1},{"id":"16405079610149095765","name":"cyberowl","vote":1},{"id":"16781982801159042389","name":"Dexter Caff/acc","vote":0},{"id":"2649066124191664356","name":"CodeGov.org (see neuron merger plan with the Wenzel known neuron)","vote":1},{"id":"4966884161088437903","name":"Wenzel","vote":1},{"id":"5553849921138062661","name":"Synapse (see neuron merger plan with the Wenzel known neuron)","vote":1}],"latest_tally":{"no":12117448824960326,"timestamp_seconds":1774325592,"total":46628482094313175,"yes":12631184166763101},"payload":{"motion_text":"This motion recommends adjusting ICP tokenomics through demand acceleration and changes to voting and node rewards."},"proposal_id":140888,"proposal_timestamp_seconds":1773762156,"proposer":"54","proposer_name":null,"reject_cost_e8s":2500000000,"reward_status":"SETTLED","self_describing_action":{"type_description":"Propose a text that can be adopted or rejected. No code is executed when a motion is adopted. An adopted motion should guide the future strategy of the Internet Computer ecosystem.","type_name":"Motion","value":{"motion_text":"This motion recommends adjusting ICP tokenomics through demand acceleration and changes to voting and node rewards."}},"settled_at":1774368000,"status":"EXECUTED","success_value":null,"summary":"\n\n### TL;DR\n\nThis motion recommends implementing Mission 70 to position the Internet Computer network, its ecosystem, and its stakeholders for long term success.\n\nMission70 targets a reduction in ICP inflation of at least 70% by the end of 2026, based on a combination of supply-side measures and demand growth. A detailed description of the suggested changes is available in the Mission70 [whitepaper](https://internetcomputer.org/whitepapers/mission70.pdf).\n\nMission70 has several components. This motion focuses on three: demand acceleration, voting rewards, and node rewards.\n\n### ICP demand acceleration\n\nMission70 relies on demand growth to increase ICP burn, driven by the following initiatives:\n\n* Powering the new \u201cself-writing cloud\u201d paradigm. A great example is the Caffeine.ai platform, which builds on the Internet Computer to provide users with self-writing cloud functionality. Already today, Caffeine.ai users are responsible for a large portion of the computation hosted on the Internet Computer, and thus for the corresponding burn of cycles and ICP. Caffeine.ai is experiencing rapid growth, reflecting how important this new trend is for the Internet Computer.\n\n* Providing \u201ccloud engine\u201d functionality. Cloud engines are for everyone, and represent the Internet Computer transitioning from a niche platform for web3 apps, to a mainstream cloud solution with unique advantages. They address a horizontal market spanning solopreneurs, startups, SMEs, corporates, NGOs and governments. Essentially, a cloud engine is a private subnet that the controller/owner (which can be a person, organization or SNS) can orchestrate and configure in various ways using a control panel. Within the rules of the network, owners can configure different replication factors, and select nodes in a single geographical region if they wish, for legal or performance purposes. They host apps that are immune to traditional cyberattacks, guaranteed to run with fault bounds, can simplify scaling, and can natively custody and process digital assets and payments. They offer a unique form of sovereignty, where the underlying compute resources are switchable: for example, a cloud engine could be created cloud-on-cloud across Amazon data centers, then switched to Google, or a mix, then onto traditional node provider hardware of the kind that powers shared subnets today, all without interrupting hosted apps. By providing a control panel based experience, cloud engines will begin to bridge the gap to mainstream cloud expectations, where customers login to websites like AWS, and see controls that allow them to orchestrate cloud resources, while maintaining decentralization. Nodes available to cloud engines do not receive any fixed payments from the network, but the nodes in an engine receive 80% of the revenues generated, while 20% are used to burn ICP, in a 100% deflationary model. Associations of node providers will be able to market cloud engines based on their nodes, benefiting from selling the Internet Computer, in a similar way that internet service providers (ISPs) benefit by creating their own internet subnets and selling the internet access, unlocking decentralized incentives and growth dynamics.\n\n* Being a \u201cplatform for agents.\u201d Increasingly, agents will build apps and systems. Sometimes they will build to directly satisfy human requests, as is the case with Caffeine.ai which spins up ensembles of agents for its human users, and other times, agents will be autonomously building systems and services for their own use, in pursuit of their aims and tasks. There are already millions of agents, and their number will one day grow into the billions. Even modern development platforms used by human developers, such as Cursor, Claude Code, Antigravity and Codex, are now powered by agents that usually do most of the actual coding. The future development of the Internet Computer shall focus on developing the network as the ultimate platform for agents to build on. For example, the simple database at https://skills.internetcomputer.org/ will soon contain 100s of skills that empower agents that are building on the Internet Computer. Furthermore, agents without traditional financial means will be given ways to begin building on ICP without friction.\n\n* Changes to \u201ccompute pricing.\u201d As discussed in previous proposals, we calculate that the network can increase the price of some aspects of onchain compute, such as subnet memory. In the case of subnet memory, where apps have substantial storage demands, such as the need to maintain large numbers of files, they will be able to migrate to the forthcoming ICP Blob Storage functionality.\n\n### Suggested voting reward changes (supply-side reduction)\n\nFor voting rewards, Mission70 includes:\n\n* Shorter dissolve delays: reduce the maximum dissolve delay from 8 years to 2 years, and the minimum dissolve delay for voting from 6 months to 2 weeks, with a one-time migration that caps existing neurons at 2 years. Reward levels would be reduced proportionally.\n\n* Convex dissolve-delay bonus curve: replace the current linear bonus with a convex quadratic function and increase the maximum dissolve-delay bonus to 3 (from 2) to keep strong long-term incentives.\n* \u201c8-year gang\u201d recognition: neurons that historically maintained an 8-year dissolve delay receive a 10% reward boost via a flag, which is proposed to remain until the end of 2030 and is lost once the neuron starts dissolving.\n* Cap the voting reward pool after bootstrapping: introduce a cap on the daily voting reward pool 8 years after genesis.\n* Simpler maturity modulation: replace the current approach with a level-based signal comparing 7-day and 365-day moving averages.\n* Impact (order of magnitude): voting rewards are estimated to decrease by about 40%.\n\n### Suggested node reward changes (supply-side reduction)\n\nFor node rewards, Mission70 includes:\n\n* Reduce legacy Gen-1 rewards: It is recommended to target an overall reduction of Gen-1 rewards by 80%, achieved through a combination of transferring existing nodes to cloud engines or offboarding them, together with reduction in Gen-1 reward levels by 40%. This recommendation is motivated by the fact that Gen-1.1 rewards substantially exceed estimated operating costs, while many Gen-1 nodes are currently underutilized.\n* Rely more on SEV-capable hardware and smaller subnets: enable smaller application subnets composed exclusively of SEV nodes, for example reducing application subnet size from 13 to 7, which would almost double available subnet capacity (while keeping cycle costs unchanged).\n\nWe should also review how interested Gen-1 node providers can obtain a clear upgrade path to newer hardware generations. In addition, Gen-1 nodes could provide low-cost infrastructure for cloud engines, which node providers could market accordingly.\n\n### Technical rollout model\n\nThis motion recommends an iterative rollout, with timing based on implementation complexity and impact assessments. Changes should be introduced step-by-step and might be refined as implementation details become concrete. For example, subnet memory pricing will be increased over a series of steps.\n\n### Additional work\n\nMission70 represents an important step in optimizing the Internet Computer\u2019s tokenomics through governance. As the ecosystem evolves, the NNS will continue refining tokenomics and related mechanisms according to the game theory that drives it, which behooves voters to adopt decisions that align with the interests of neuron holders over their neuron dissolve horizons (i.e. if a neuron can be dissolved in 2 years, what will be the value of the ICP at that future moment).\n\nTo prevent ICP tokens hitting public markets in an uncontrolled way, and to provide global markets with the maximum possible confidence in ICP, as mentioned in the #Mission70 white paper, the NNS can introduce a dissolution/disbursement queue similar in spirit to the Ethereum validator queue, but make it far more adaptive to market conditions. If the amount of ICP scheduled to dissolve on a given day exceeds a governance-defined threshold, which the NNS can dynamically adapt according to prevailing market conditions, the queue could automatically stagger disbursements from dissolutions over a forward period.\n\n","title":"Mission70: Demand acceleration and  adjustments for voting rewards and node rewards","topic":"TOPIC_GOVERNANCE","total_potential_voting_power":53020082586350733,"updated_at":"2026-03-24T16:00:50.114394","url":"https://forum.dfinity.org/t/mission70-demand-acceleration-and-adjustments-for-voting-rewards-and-node-rewards/65322"}