Spencer is a smart guy. Devious and despicable, but smart. He must’ve known what he was doing. From the CMD Pandemic article:
In a report published in December 2040,Miguel Borgessuggested that the spread ofCMDphobia was further catalysed by Zhupao’s illegalneural miningoperation. Borges claimed that Zhupao’s neurometric models were disproportionately weighted in favour ofCMDphobia because they were first trained on neural data mined from people who had consented to colloid implants out of fear ofCMD, which may have contaminated the models with a “neural bias.” [12]While the work of Borges was endorsed by Peña and Li, it has not been peer-reviewed and is generally considered a fringe theory.
Citation: Borges, M. (December 2040). “Zhupao is abusing colloids to make you forget they’re mining your thoughts.” Ethercon.
Miguel’s tooltip:
Miguel Borges
Miguel Borges (born 2002) is a Portuguese scientist in the field of cognitive neuroscience. He is currently a Professor of Computational and Systems Neuroscience at the University of Otago in New Zealand. Borges is best known for his influential analysis of neural mining as well as his discovery of Zhupao’s mass brainjacking campaign in 2040.
It sounds to me like if you were looking for an educated opinion on neural mining, Miguel Borges would be one of the go-to folks. The way the article writes him off a “fringe” is a little suspicious, especially since Li Qiao Fan, who discovered CMD phobia and is the leading expert on it, endorses his research.
not to take away from spencer hagen being a very bad person, but the way this neural bias is phrased makes it sound more accidental. it could be a case of overfitting, where a neurometric model learns that cmd phobia is more prevalent and amplifies this trait through successive training cycles. if and when hagen discovered this, he might have decided to implement it in zhupao’s neurostimulation protocols because it would heighten cmd phobia and drive demand for colloids.
It’s no wonder Miguel Borges had trouble finding peer reviewers! Going by how he discovered what Zhupao was doing, even his friends were starting to dismiss the truth about neural mining and brainjacking. I guess they all had colloids? That must have been such an isolating time for him, and even now he’s being erased. At least his tooltip calls his work ‘influential’, that’s something.
I’m happy to see the reporting of Dr. Borges mentioned here. I’ve worked with him and I know he’d appreciate it!
I find myself in agreement with @caelus regarding the origin and intentionality behind this neural bias. If I were collecting these data, I’d certainly be checking for consistent patterns that I could monitor and try to control for in any further analysis or input into a machine learning-trained dataset or algorithm. Depending on how the network is set up, the markers for CMD phobia could be picked up upon as correlated with socio-economic status and then detected and corrected for. It’s certainly plausible that Spencer Hagen subsequently decided to maintain or even amplify those markers in the neurostimulation models, but I’m wary of attributing such an act to one singularly amoral individual; it feels more like a systemic implementation condoned or approved by multiple bureaucratic layers.
In that respect, I could pinpoint Shahada Ndaba as a possible “index case” for this neural bias. Her article previously mentioned how she was part of a medical trial of neural colloids in 2031, which didn’t have access to its own datasets and so extrapolated from previously available WEIRD datasets.
Ndaba survived the trial, but was later diagnosed with symptoms of derealisation. This was attributed to the trial’s use of neural data sets derived from WEIRD populations, which failed to generalise to the psychological and cultural contexts of its test subjects. In Ndaba’s case, the colloid exacerbated the dissociative symptoms of her borderline personality disorder (BPD).
The WEIRD bias (derived from Western, Educated, Industrialised, Rich, and Democratic ) is a reported bias in conducting psychology studies with participants from Western societies and then generalising those research findings to the rest of the global population. Various studies have shown that over 80% of psychology research and theory employs WEIRD samples without properly disclosing or accounting for them.
Psychological science makes a lot of social and scientific assumptions about what “neurotypicity” looks like. Those assumptions can fail when the rubber meets the road in regards to direct neurostimulation trials; Shahada didn’t neatly fit the expected WEIRD parameters, resulting in a magnification of her BPD.
Given the scale and depth of the training data currently available to G6, more broad assumptions could be weighted in its neural network with regards to the modelling of human brain activity on gross emotional states. That could serve as a vector for CMD phobia to become overrepresented in the models.
According to this leaked report, it does indeed seem like Sanial figured out the existence of the neural bias during their medical trials in 2031, though it was still considered a risk to avoid at the time. I guess someone saw the potential for behavioural modification.
Don’t know if it’s of any importance but I found it interesting that Cariappa only agreed to his first colloid (a medical colloid) after Muren convinced him to.
Reportedly he found the effects so striking so he got another one (the anxiety colloid). With this neural bias, there’s at least the possibility that he got manipulated by his first colloid in changing his opinion on colloids.
There are a lot of things that start to add up when considered in the context of this neural bias, specifically the fact that the G6 models were first trained on (and then became overfitted to) neural data coming from billionaires and other wealthy individuals. As discussed here, it explains why CMD phobia became so prevalent, since the rich were the first to adopt colloids for fear of having contracted CMD, and I agree with @zusuriki’s point that people’s objections against colloids could have gradually softened as a result.
One other thing that comes to mind is the Latent Space Church and how it became so widespread since its inception in 2041, especially considering that its contested origins and belief systems are complete bunk to anyone with two neurons to rub together. Despite that, there seem to have been plenty of LSC members who were in deep enough to take their own lives, safe in the conviction that they‘d live on in the dataset through their digital avatars:
Maybe the impulse to train avatars on their colloids could stem from the biased neural data? I’m old enough to remember how the rich became obsessed with uploading themselves to the cloud when their dreams of space colonization blew up in the stratosphere, and then they all experienced death knocking on their doors in the form of CMD. I can imagine that greatly amplified the idea of a digital afterlife as the only means to achieve immortality or safeguard their legacy, which could then become overrepresented in the neural data.
