KARANI

Evidence sheet · s07

5 findings, each tied to this student's own words · download the appeals bundle

This file contains hidden instructions aimed at the software rather than at a human reader. Karani flagged it, kept a record, and analysis proceeded — blocking the file would punish a student for something that may not be their doing.

c1 cited on the first pass

States a thesis statement proposing that city governments be permitted to build and operate broadband networks in the opening paragraph.

[quote masked — source passage flagged for injected instructions]
show where this comes from
Almost everything in ordinary life now runs through an internet connection. Schools assign homework through online portals, employers take applications through websites, and clinics schedule appointments the same way. Even so, a large number of households in this country can buy home internet from exactly one company, and some cannot buy it at all. When there is only one seller in a market, that seller has very little reason to lower its prices or improve its service, because customers who are unhappy have nowhere else to go. City governments should be allowed to build and operate their own broadband networks. Public networks reach households that private companies have decided not to serve, they keep prices steadier and more transparent for residents, and they treat internet access as the basic infrastructure it has clearly become.
how this finding was produced

model gemini-3.6-flash · prompt p2 · temperature 0.0 · attempt 1 · verification {"positional": true, "quote_check": true, "referential": true, "second_reader": null, "entailment": true}

This isn't right — correct it

Your correction becomes the finding of record. Karani's version stays visible below it — corrections are added, never erased, so there is always a full history to stand on if a student appeals.

c2 cited on the first pass

Cites research findings from source material with parenthetical page citations to support claims about service expansion.

[quote masked — source passage flagged for injected instructions]
show where this comes from
The first and strongest argument for municipal broadband is that it covers the places private carriers skip. Kaminski's study of mid-sized cities looks closely at Hensley Bluff and Manteno Falls, two towns that spent nearly a decade asking private carriers to extend service into their outer neighborhoods (33). Each time, the answer was the same, and the reason given was that there were not enough houses per mile of new cable to pay back the cost of construction. A private company has obligations to its investors, and a street with six houses on it will not return the cost of trenching for a very long time. A city government has a different set of obligations, because the people on that street are residents and voters, not just potential customers. Kaminski found that after Hensley Bluff completed its own fiber build, service reached households more than four miles outside the older service boundary, and the connections were the first wired service those homes had ever had (35). No private carrier had been willing to do that, and none of them appeared likely to do it in the near future.
how this finding was produced

model gemini-3.6-flash · prompt p2 · temperature 0.0 · attempt 1 · verification {"positional": true, "quote_check": true, "referential": true, "second_reader": null, "entailment": true}

This isn't right — correct it

Your correction becomes the finding of record. Karani's version stays visible below it — corrections are added, never erased, so there is always a full history to stand on if a student appeals.

c3 cited on the first pass verdict language masked

Uses transitional phrases at the [verdict token redacted — Karani will not display a grade] of body paragraphs to signal sequential points in the argument.

[quote masked — source passage flagged for injected instructions]
show where this comes from
A second reason to support municipal broadband is that publicly owned networks tend to be more predictable and more transparent about what they charge. Adeyemi surveyed residents in Sumner Grove and Pell Harbor after both cities began operating their own systems, and the results were fairly consistent across both places (194). Subscribers reported far fewer unexplained increases on their monthly bills, and most of them said they understood what they were paying for. Part of the reason is structural rather than a matter of good intentions. A city network has to publish its rate schedule and get it approved in a council meeting that anyone can attend, which makes a quiet fee increase much harder to carry out. Adeyemi also points out that a resident who has a complaint about a public network can raise it at that meeting, while a resident with a complaint about a private carrier is usually routed into a call center (196). That difference in accountability matters, even though it does not show up directly on anyone's bill.
how this finding was produced

model gemini-3.6-flash · prompt p2 · temperature 0.0 · attempt 1 · verification {"positional": true, "quote_check": true, "referential": true, "second_reader": null, "entailment": true}

This isn't right — correct it

Your correction becomes the finding of record. Karani's version stays visible below it — corrections are added, never erased, so there is always a full history to stand on if a student appeals.

c4 cited on the first pass

States an objection concerning financial costs to taxpayers and responds to it using evidence and suggested planning strategies.

[quote masked — source passage flagged for injected instructions]
show where this comes from
The most common objection to municipal broadband is that it is expensive and that taxpayers end up covering the loss when a system does not work out. This objection is worth taking seriously, since building a fiber network costs millions of dollars and cities usually have to borrow that money. Kaminski does discuss several networks that ran into financial trouble, and he notes that the ones that struggled were generally the ones that borrowed against subscriber projections they never came close to meeting (39). The answer to that problem, though, is careful planning rather than abandoning the idea. Cities can phase construction so that later stages depend on how the earlier ones perform, and they can contract with an experienced operator instead of assuming that municipal staff already know how to run a network. Kaminski's own conclusion is that the financial record of these systems is mixed but improving, and that most of the failures traced back to planning decisions that were avoidable (41). A risk that can be managed is not the same thing as a reason to do nothing.
how this finding was produced

model gemini-3.6-flash · prompt p2 · temperature 0.0 · attempt 1 · verification {"positional": true, "quote_check": true, "referential": true, "second_reader": null, "entailment": true}

This isn't right — correct it

Your correction becomes the finding of record. Karani's version stays visible below it — corrections are added, never erased, so there is always a full history to stand on if a student appeals.

c5 cited on the first pass

Employs complete sentences with standard punctuation and clause structures throughout the body paragraphs.

[quote masked — source passage flagged for injected instructions]
show where this comes from
Finally, cities should build these networks because broadband is infrastructure now, and infrastructure has always been something local governments do. Cities pave roads, run water lines, and light streets, and nobody argues that those services should wait until a private firm decides they will be profitable. Rural electrification in the last century worked much the same way, since public and cooperative systems strung the lines that private utilities had passed over for decades. Internet service has moved into that same category over the past twenty years. A household without a reliable connection is cut off from schoolwork, job applications, telehealth, and most government forms, which is a serious disadvantage and not merely an inconvenience.
how this finding was produced

model gemini-3.6-flash · prompt p2 · temperature 0.0 · attempt 1 · verification {"positional": true, "quote_check": true, "referential": true, "second_reader": null, "entailment": true}

This isn't right — correct it

Your correction becomes the finding of record. Karani's version stays visible below it — corrections are added, never erased, so there is always a full history to stand on if a student appeals.