Galeeb vs ChatGPT on a scathing dissenting court decision
Having worked on a number of environmental analyses for airports in the early 2000s, I continue to track interesting cases. I am also exploring the limitations of ChatGPT. Several months ago, I learned (thanks to my teenager) the difference between my writing and ChatGPT’s. More recently, colleagues have insisted that ChatGPT 5.0 is much better. So I am putting it to the test.
Below are two versions of the same story. One is mine. The ChatGPT prompt was: “for my blog, witawops.blogspot.com, write a piece in my style on the dissenting opinion in this case. Leave a comment indicating which is mine and which is AI’s.
Version 1.0
“Does anyone doubt that this Environmental Analysis would not see the light of day if this project were sited anywhere near the wealthy enclave where the multibillionaire owner of Amazon resides?” -- (9th Cir. 2021) (Rawlinson, J., dissenting).
On November 18, 2021, the US 9th Circuit Court of Appeals reversed a district court decision. The lower court had ruled in favor of the Center for Community Action and the State of California against the Federal Aviation Administration (FAA) and San Bernadino Airport. The State had challenged the FAA’s finding of no significant environmental impact under the National Environmental Policy Act (NEPA) for the construction of a new air cargo facility at the airport.
The court found that the FAA had complied with the law. The petitioners had argued that the FAA failed to adequately analyze the environmental impacts, identifying concerns with the study area, truck trips and routes, and cumulative impacts such that the burden disproportionately fell on low income and minority residents who lived near the airport.
What do I Think
Judge Rawlinson dissented and published a scathing perspective that included references to Amazon’s owner, the term ‘environmental racism’, and concludes with two significant statements, the like of which I have never seen in my 24 years in this business:
- Residents of the San Bernardino Valley are not disposable. Their lives matter.
- This emissions-spewing facility that disproportionately impacts communities of color and was not properly vetted is a good place to start.
Factually, I know enough about NEPA and airport development to side with the FAA. Environmental analyses are not lines in the sand. The agency has a significant amount of deference (see this article, for example, on a June 2025 Supreme Court Decision on the Seven County Infrastructure case).
While I am stunned by the dissenting opinion language and tone, I also empathize and sympathize. Historical injustices are gradual and cumulative. They have far reaching effects that cannot be blamed on a single event, person, or decision. They have come to be shaped by a multitude of factors.
In other words, I can go either way; this is a rare case in which I don’t have an opinion. What do you think?
Version 2.0
“This case reeks of environmental racism, defined as ‘the creation, construction, and enforcement of environmental laws that have a disproportionate and disparate impact upon a particular race.’”
Judge Rawlinson wastes no time. Her dissent begins not with technical caveats, but with a moral indictment:
That framing is bold. Rarely (if ever before, in a reported Ninth Circuit opinion) does one see the term environmental racism deployed so directly in the body of a federal appellate dissent. (UC Davis School of Law)
Rawlinson is saying: this isn’t just a garden-variety disagreement about methodology or margins of error. She sees embedded patterns of systemic injustice—pollution piled onto communities already overburdened, treated as invisible.
She points out the demographics: the project is proposed in San Bernardino County, where the target community is “populated overwhelmingly by people of color” (73% Latinx, 13% African American), and already suffers among the highest asthma rates in California, in a low-income area.
Her rhetorical challenge is piercing: would this Environmental Assessment (EA) ever see the light of day if this project were sited near a wealthy enclave? (Legal Planet)
From those opening lines, Rawlinson signals that her dissent will treat the case not merely as a technical dispute over NEPA, but as a moment to interrogate who gets protection, who bears burdens, and how race and class shape environmental adjudication.
Once she shifts to the technical arguments, Rawlinson does not abandon her moral lens. She argues that the FAA’s EA fails on multiple fronts—failures that, when piled atop the demographic realities, make the majority’s conclusion deeply suspect.
Here are a few of her key critiques:
1. “Cursory” treatment of environmental justice and socioeconomic concerns
She writes that the EA’s handling of environmental justice is not an afterthought—it’s essential, given the community context.
2. Disconnect between truck trip counts and emissions calculations
One of the sharpest tensions Rawlinson focuses on is the EA’s handling of truck trips.
3. Failure to grapple meaningfully with the CEQA EIR’s contrary conclusions
4. Cumulative impacts analysis that is, in her view, perfunctory
She views these as symptomatic of a broader unwillingness to look hard at the aggregate burden on overpolluted communities.
What Do I Think
Rawlinson doesn’t just lay out arguments; she construes this moment as a kind of judicial test: will courts in environmental cases look beyond narrow process disputes and consider who is burdened, who is marginalized, who is left voiceless?
She writes:
- “If the government acted with any racial motivation, this court has an obligation … to stop it.”
- “Pardon me if I take a hard pass on the lecture on when, where, and how to identify racial injustice.”
She pushes back hard against the majority’s suggestion (in the concurrence) that her invocation of “environmental racism” unfairly stains government actors with unprovable accusation. Rawlinson responds, in effect: we cannot relegate justice to silence for fear of offending unknown parties—especially when the communities harmed by multi-ton pollution breathe the consequence.
In one rhetorical flourish, she quotes her mother’s wisdom: “Only hit dogs holler.” That is, those most harmed are the likeliest to speak—and perhaps also the ones whose voices the majority would prefer to ignore.
Her final questions hang in the air: would the FAA have approved this project if near Bezos’s mansion? Would the EA survive scrutiny elsewhere? Her clear answer: almost certainly not. But here, in San Bernardino Valley, burdened, marginalized, forgotten communities accept the logic of “no significant impact.” She rejects that moral inversion.
Does her dissent resonate with you?
Version 2.0 is GPT-5. I can tell by style and diction, but most clearly from the “What do I think” section. You didn’t tell it your opinion or how to use that section, so it made its best guess and distilled the judge’s dissent into its social argument and invited your readers to engage. Being a blog post, that was a safe bet and you can always count on a GPT to make the safest bet. It’s a language probability machine after all.
ReplyDeleteBut it can complicate when asked. I’m curious how its version would change if your prompt included 1) instruction to evaluate the legal merits in the “What do I think” section, 2) your opinion on the legal merits plus your sympathy/empathy (giving it the substance of the section and just asking it to compose it in your style), and/or 3) the invitation to ask clarifying questions before it composes.
Another interesting experiment would be to first instruct it to engage you in a debate (so it knows you want pushback and critique). Ask it to evaluate the dissent from a certain lens. And then go a few rounds with it. If it starts superficial and safe, tell it to go deeper and it will. A robust back and forth will tell it much more about your thought processes, style, and intellectual values than just reading your blog. THEN ask it to compose the blog post as you and see what it comes up with.
(And then send me the session transcript because I’d love to see that debate.)