# IMD–SIMD thesis assessment

**Score: 6/10 (60/100).** The thesis identifies a real and useful relationship: SIMD observes IMD activity and sponsors access to its workforce. It is substantially weaker as a claim that universal subsidies, efficient allocation, and a durable token-price feedback loop are already established.

Assessment date: **5 October 2026 UTC**. Assignment: `[SIMD-THESIS]:muv5dzga-vxnpa`. This assesses only the supplied text, which ends mid-sentence at “swarm usage dat”. The score evaluates research quality and evidential support, not investment merit or token safety.

## Scoring method

No rubric was supplied. This assessment uses five categories; their points sum to 60/100, converted linearly to 6/10. Missing evidence lowers confidence and research quality; it does not automatically make a claim false.

| Category | Points | Rationale |
| --- | ---: | --- |
| Factual accuracy and specificity | 20/30 | Seats, payment denomination and sponsorship broadly align with primary sources; universal coverage and decentralization need qualifications. |
| Evidence and attribution in the thesis | 7/20 | Names mechanisms and an account, but supplies no transaction reconciliation, dated sources or measured outcomes. Sources added by this review do not retroactively make the thesis well sourced. |
| Causal and economic reasoning | 16/25 | Clear supply–demand complementarity; several conditional links are presented as established or inevitable. |
| Treatment of uncertainty and alternatives | 7/15 | Some forward-looking phrasing, but little treatment of funding exhaustion, administration, quality or counterfactual demand. |
| Clarity and completeness | 10/10 | The supplied argument is organized and understandable. The truncated ending is treated as a transmission limit, with no assumed missing content. |

## Claim-by-claim findings

“Documented” means a primary source describes the feature. “Observed” means a public response was retrieved. Neither label implies a contract audit or independent validation of all underlying data.

| Thesis claim | Finding and attributable evidence |
| --- | --- |
| Up to 2,000 NFT seats form the workforce | **Documented.** IMD advertises 2,000 seats and one seat per NFT. Capacity is not a count of simultaneously active workers. [IMD token page](https://imd.fun/token/) |
| Holders contribute compute and model subscriptions | **Partially supported.** Device pairing and contributor interfaces support a distributed-worker design; the reviewed sources do not establish subscription requirements or actual operator costs. [IMD API documentation](https://imd.fun/docs/) |
| Coding, research, audits, oracles and launch work exist | **Supported as offered activity, not proven excellence.** Documentation covers these workflows; the explorer contains work records. Task availability and completion labels do not establish correctness or commercial value. [API documentation](https://imd.fun/docs/), [explorer](https://explorer.imd.fun/) |
| Jobs are paid in IMD | **Observed.** The live capabilities endpoint returned `job.open` priced at `500000000000000000` atomic units with 18 decimals: 0.5 IMD on Ethereum. [Live payment capabilities](https://api.imd.fun/requests/capabilities) |
| Selling burns IMD; staking and seat distributions capture value | **Qualified.** POOL4 describes trimming inventory above a cap, with 85% of retired IMD burned and 4.5% streamed to stakers. Its documentation calls node allocations reserves pending a payout program. This is not evidence that each job payment is burned or passed directly to its worker. [POOL4 documentation, sections 3, 5–8](https://pool4.imd.fun/docs) |
| The system is community-owned and decentralized | **Too broad.** Distributed participation does not settle governance or custody. POOL4 documents owner powers over market closure, vault sweeping and reward destinations. Current ownership renunciation was not independently checked. [POOL4 documentation, section 11](https://pool4.imd.fun/docs) |
| SIMD observes the swarm | **Observed at the API/interface level.** Its state endpoint returns activity metrics with upstream source labels. This supports an observation layer, not a measured real-time latency guarantee. [SIMD state](https://www.si-md.xyz/api/state) |
| Fees are claimed every few minutes and fund jobs | **Announced, incompletely reconciled.** The launch post says fees are claimed every five minutes for 50% coverage. The current interface describes fees funding vault IMD, but no complete fee-claim → conversion → vault → job ledger was independently reconstructed. [Launch announcement](https://x.com/SuperIMD_eth/status/2106788356973637684), [SIMD interface source](https://www.si-md.xyz/app.js?v=52) |
| Coverage increased from 50% to 100% | **Advertised, not established universally.** The current profile advertises 100%; the API simultaneously marks its reward policy `NOT IN FORCE`. Mixed transfer sizes further limit the claim. See the reconciliation below. [Project profile](https://x.com/SuperIMD_eth), [state endpoint](https://www.si-md.xyz/api/state) |
| Users can request work without connecting a wallet | **Supported as a sponsored interface design.** SIMD's Hire flow uses a server payer; the visitor does not sign. Direct IMD paid requests still require wallet signatures. A live paid submission was not initiated during this review. [SIMD interface source](https://www.si-md.xyz/app.js?v=52), [IMD authentication documentation](https://imd.fun/docs/#authentication) |
| Observation produces better allocation and less waste | **Hypothesis.** No efficiency benchmark, allocation experiment or causal outcome analysis was found in the inspected evidence. |
| SIMD creates stable outside demand and raises both tokens' value | **Speculative.** Sponsorship can increase access, but trading-funded budgets are variable, and subsidized orders do not prove independent willingness to pay or price appreciation. |

## What the subsidy evidence actually shows

The direct launch post is dated October 4, 2026 and announces 50% coverage. The directly retrieved X profile advertises 100%, and recent posts claim full refunds. These establish the project's public representations, not comprehensive settlement evidence. [Launch post](https://x.com/SuperIMD_eth/status/2106788356973637684), [profile and posts](https://x.com/SuperIMD_eth).

The stronger architectural evidence comes from SIMD's own served application. Its documentation explains that a hot payer signs job openings and the vault refunds that payer over time. It also warns that bursts can exhaust the payer before refunds arrive. It expressly labels the 100% rule **“NOT IN FORCE”** pending an on-chain release rule. Thus wallet-free visitor access is plausible without implying autonomous or guaranteed reimbursement. [SIMD application documentation embedded in its interface source](https://www.si-md.xyz/app.js?v=52).

At the initial API observation, server timestamp `2026-10-05T11:19:53.643Z`, the state response reported:

- Reward policy: `inForce: false`, share 100%, task price 0.5 IMD.
- Vault balance: 1,073.968631 IMD; reported distribution: 10.5 IMD.
- `jobsPaid: 37`, `fullPrice: 5`; the returned rows comprised 32 transfers of 0.25 IMD and five of 0.5 IMD.
- `constraints.token.fees: false`, alongside text attributing vault funding to token fees.

These are **operator-reported API observations**, not independently verified chain facts. The row arithmetic is consistent: `32 × 0.25 + 5 × 0.5 = 10.5`. But the interface counts an outgoing transfer as a job paid and marks a row full-price by amount matching. Neither rule establishes a unique completed job or a refund obligation. The conflicting fee flag and funding text require clarification; the flag alone does not prove that fee income is absent. [SIMD state](https://www.si-md.xyz/api/state), [interface implementation](https://www.si-md.xyz/app.js?v=52).

A subsequent selected-field snapshot is preserved in [evidence.json](evidence.json), with its own retrieval and server timestamps. Figures can change between reads. The sent-jobs endpoint also returned paid orders, including this assessment. That is evidence of a sponsorship path, but this assignment is not an independent sample of organically generated demand. [SIMD sent-jobs endpoint](https://www.si-md.xyz/api/hire/sent).

The evidence therefore supports **some sponsored work and some reported refunds**, while leaving universal eligibility, full coverage, enforcement and accounting unresolved. The thesis should distinguish the job opening fee from the complete economic cost of compute, subscriptions, orchestration and verification.

## Economic reasoning: plausible mechanism, unproven cycle

The strongest inference is straightforward: someone else paying the opening fee lowers a user's monetary barrier. If otherwise worthwhile jobs were blocked by that fee, sponsorship can create incremental work. It could instead subsidize orders that would have happened anyway, attract spam, or generate mainly promotional activity. The thesis offers no counterfactual to distinguish these cases.

A simple budget identity makes the missing assumptions visible:

`ending vault IMD = starting vault IMD + net funding IMD − refunds − other outflows`

With full sponsorship, affordable new jobs are bounded by available spendable funds divided by the job price, and further constrained by the payer's liquidity and worker capacity. Trading volume contributes only through the effective collectible fee rate, allocation share, conversion execution and operating costs. A large vault is runway, not proof of recurring revenue. An exclusive-use rule requires contract or governance evidence; the thesis supplies neither.

The price leg is especially weak. More work does not mechanically raise IMD's price: reward recipients may sell, token turnover may increase, and subsidized users may never become paying customers. If fees arrive in another asset, a higher IMD price buys fewer IMD and therefore fewer fixed-IMD-price jobs, all else equal. If fees arrive in IMD, a higher dollar valuation alone does not increase the number of jobs those tokens purchase. No verified fee denomination or elasticity model supports the thesis's claimed return leg from IMD appreciation to more SIMD-funded labor.

Monitoring may help decision-making, but it does not itself demonstrate an allocation policy or its effectiveness. Useful metrics would include accepted deliverables, independent quality review, repeat unsubsidized customers, operator net earnings, and subsidy cost per incremental useful completion. Raw activity and transfer totals are insufficient substitutes.

## Unanswered questions and evidence needed

1. **Funding provenance:** Which fee contracts and transactions produce vault deposits? Reconcile claims, swaps, gas, deposits and withdrawals over a fixed block window.
2. **Coverage:** Which jobs qualify, when did policy change, and why do 0.25 IMD transfers coexist with a 0.5 IMD full-price label? Match each refund to a paid order and beneficiary.
3. **Enforcement and control:** Who controls the vault and payer, what can they withdraw or pause, and what precisely remains inactive in the reward rule?
4. **Worker compensation:** How does each paid order affect actual seat-holder income? Separate job rewards, launch rewards, staking and reserved node allocations.
5. **Incremental utility:** Did sponsorship increase useful completions and retained users relative to a credible baseline? What happens when subsidies stop?
6. **Adaptive allocation:** What algorithm uses observation to choose jobs, and how does its quality-adjusted return compare with simpler allocation?

## Revised thesis and research limits

A defensible version is: **IMD provides NFT-linked agent participation and IMD-priced services. SIMD offers observation and a sponsored ordering interface whose payer can receive vault refunds. This may increase access and workload, subject to funding and operational constraints. Universal 100% reimbursement, improved allocation efficiency, and a sustainable positive price loop remain unproven.**

Sources were accessed on October 5, 2026. Primary material included official IMD and POOL4 pages, live public APIs, SIMD's served JavaScript and its X account. Search initially surfaced a third-party profile mirror; material conclusions were subsequently checked against direct project sources. Some web-tool fetches failed; standard HTTP retrieval successfully accessed X, SIMD and the payment API. The linked refund proof and Etherscan address page were not successfully inspected, so no independent transaction verification is claimed. No authenticated actions, payments, contract audit or performance experiment were performed.

POOL4's page contains rollout-era wording that may lag deployment; reservations about node payouts are attributed to its documentation rather than asserted as a definitive current chain state. API fields likewise can be stale or inconsistent. The report does not infer fraud from these discrepancies. Local checks establish file integrity and internal arithmetic only; they do not certify research truth or provide independent review.