Daniel Coimbra

AI programmer for a living, academic philosopher by vocation. Passionate about foundational issues in logic, comp, maths, stats, physics, bio, neuro, and AI!

[Daniel Coimbra @dan_sci_phil]

Opinions have three properties: content strength, confidence level, and affective stance.

Content can be strong or weak. Confidence can be high or low. Affective stance can be arrogant or humble. These are independent, except that being unsure entails being humble.

Consider an opinion about telekinesis. I have provided a table of possible opinion proprerties below.

Two observations:
1. Most people associate a strong opinion with an arrogant opinion. This is a mistake, as the table shows.

2.Most people associate a weak opinion (such as agnosticism) as humble. This is also a mistake! Agnosticism can be arrogant if it considers non-agnostics to be presumptuous.

[Daniel Coimbra @dan_sci_phil]

So far we have debated the merit of a thesis.

You propose we lift the subject to focus on our psychological motivations. I accept.

Thirty years ago, I was born into a secular family in southeastern urban Brazil. My teenage self underwent an ideological reaction against religious conservatism in Brazil and in the US (to which I had internet exposure). This prompted me to research arguments against my "target".

Because of this, by the time I became 16 my intellectual development was already deeply entrenched with scientism, physicalism, naturalism, analytic philosophy, and Enlightenment epistemic values. Like almost every other people, I was a machine of motivated reasoning.

From there onwards I have attempted to be rational, self-aware, tolerant, and intellectually honest. My greatest method is to argue rationally, without affection or rhetoric, so that my opponent has as much surface as possible to expose my reasoning flaws. Naturally, increasing success at this came with age, even if no human can ever claim utter success.

I can however attest that I have made extensive efforts to find reasons to change my views, many of which I did change. Nonetheless I have never found reason to believe the paranormal. In fact, I see systematic reason to deny it. From all my self-understanding efforts, I do not gather that this has an emotional cause. If I am wrong, the cause is cognitive limitation and lack of information.

I close with two remarks.

[Remark #1]
As you say, there may be uncontrolled unknown variables outside physics. But we could only know this if their effects on our sense organs had minimal clarity, coherence, and stability. I mean *minimal*, since a 55% effect with p < 0.05 can already count as evidence.

Such a minimal effect is a precondition for any ordinary person to have had personal experiences which they can have good reason to believe are truly paranormal. The thing is: this same minimal effect automatically enables the paranomal to be scientifically detected.

So either (A) both scientists and ordinary people have access to the paranormal, however subtle and circumstantial, or (B) neither scientists nor ordinary people have any such access, since the paranormal is massively irregular.

In case A, absence of experimental evidence *is* evidence of ontological absence; if it was there, we should have seen it; if people saw it, scientists should see it. In case B, we have zero evidence, and the rational stance then is to default to our current systematic view of reality as purely physical.

[Remark #2]
The differences in our attitude shown during this exchange suggest that your affective attachment to your views is much greater than mine; such attachment would seem to be the reason why you have lifted the debate into a psychological accusation.

[Daniel Coimbra @dan_sci_phil]

Thanks for the info. The Wiki article you have sent is not favorable to remote sensing, but this could be an ommission.

The US Military engaged in a decades-long Project Stargate to find military applications of parapsychology. As it seems, it was declassified precisely because no useful military application was found.

Source: https://t.co/hQ6HYC7fIM

There *was* a slight effect where vague general characteristics were recovered via remote sensing, but it was so vague and inconsistent as not to be actionable. This makes me think that such vague recovery simply harvested the ordinary predictive ability of military personnel.

Your veteran acquaintance might have felt convinced, but I argue that this corresponds to his personal inclinations (and susceptibility to confirmation bias over coincidences) rather than significant evidence.

[Daniel Coimbra @dan_sci_phil]

@Dr_L_Strickland @OUPAcademic Wow! Thanks for this contribution!! He is an inspiring thinker, and it's one of the modern philosophers I know the least about.

[Daniel Coimbra @dan_sci_phil]

Reproducibility, simpliciter, does assume the causal effect is 100% present. Reproducibility with statistical significance does not. This is precisely the reason why statistical analysis is required. Individual experiments can never control for every relevant variable.

Naturally, the paranormal realm would contain an exceptional number of unknown variables, which are therefore uncontrolled for. However, if anyone on Earth has any rational reason to believe the paranormal realm exists, it is because they are picking up on statistical signals, and therefore experimenters could pick up on it. If the paranormal were irregular enough to be experimentally undetectable, then it would not be detectable by humans in ordinary life as well, so that would be a Pyrrhic victory for paranormal enthusiasts.

Your very reference to Sheldrake's studies is an instance of reproducibility with statistical significance.

Setting aside these methodological considerations, about which perhaps we can now see that we are in agreement, we could proceed to discuss scopaesthasia.

I am not sure we could get anywhere. I asked GPT 5.5 High to review current literature. It seems that any positive result is very mild, inconsistent between experimental designs, and methodologically contested. If there is a real effect, *I* cannot see it by perusing the internet from my office chair. If you have something you would find interesting to tell me, I'm all ears.

[Daniel Coimbra @dan_sci_phil]

@PaulRRobichaud I am ignorant of the policy details. I am worried about Tiktok-like brain rot. It is a more pressing problem than teen loneliness. So I'd like to see a balanced regulation where teens can participe in forums and watch long-form content, but not be subjected to reels / shorts.

[Daniel Coimbra @dan_sci_phil]

I see. We must then think about what grounds we have to believe such phenomena exist. While certain personal experiences are strong and convincing, I believe they are best explained as endogenous neurological activity. The litmus test is reproducibility with statistical significance.

Scopaesthesia for instance is simple to test experimentally, and no effect has been shown. This shows how biased human cognition is in dealing with coincidences.

There is a probabilistic argument against such paranormality: if it existed, it would very likely produce strong experimental signals, which it does not.

[Daniel Coimbra @dan_sci_phil]

Any ontological claim entails an epistemic claim, so I am not sure whether you are accusing me of something. If reality is indeed causal patterns, then by definition causal patterns are sufficient to explain all of reality.

This does not mean it will always be easy, knowable, cognitively tractable, or practical.

The real debate is: what grounds do we have to believe that reality is one way or another?

It is only then that an epistemic claim comes in as substantial information: Theories based on causal patterns have worked systematically well, all the way from particle physics to neuropsychology. Other theories have not achieved much (e.g. aristotelianism does nothing for biology, dualism does nothing for psychology). Therefore, the causal-function view is by far the most well-supported. Therefore, it should be our highest-credence view. People can rationally form alternative hypotheses, so long as they recognize that they are betting on a view that is currently under-supported.

[Daniel Coimbra @dan_sci_phil]

Oh. Sure. Any position about the mind-body problem is a strong ontological thesis. This is not the issue.

My thesis is not that reality can be known fully. That would be an epistemic claim, not an ontological one.

My thesis is offered as a most likely or coherent explanation: the mind, like everything else, is a physical causal system from top to bottom.

[Daniel Coimbra @dan_sci_phil]

@PhilosophyOfPhy Perhaps the AI writing has been overdone?

[Daniel Coimbra @dan_sci_phil]

I do happen to think that TV has been a net negative technology!

While I recognize questions such as these are enormous in complexity, I am humbly in favor of far greater restrictions on the hedonistic impulses of the young.

This debate pulls a thread that goes all the way back to the 1960s cultural wars between different visions for civilization: `authority vs autonomy`, with respect to family, culture, social expectations, academic performance, sexual behavior, entertainment, and so on.

[Daniel Coimbra @dan_sci_phil]

I think it is important to get clear on different scenarios. You're clearly making an important point that I don't want hidden behind communication noise.

Ordinary individuals assessing what is plausible by brainstorming with (or guillibly querying) AIs might be an improvement over the previous norm: to think incompetently by oneself, to trust your smartest-sounding friend, to absorb social media echo chambers, to check tabloid newspapers, etc.

Having made this point, which is something I have been thinking, I want to make clear that I understand that it is separate from other concerns you have raised. There is the problem of a hidden corporate/LLM agenda. There is a long-term problem of what happens if we get ASI. There is the problem of us nurturing even less our independent thinking abilities. And so on. All real problems!

[Daniel Coimbra @dan_sci_phil]

That is the standard view. I convinced myself that we can. Our judgments are the sole authority about whether there are qualia such as red colors and sharp pains. I am fine with an error theory that explains why it seems so compelling that such qualia exist. I believe you should open yourself to the possibility of thinking how this could be the case.

Qualia are conceptually incomprehensible objects. It makes no sense how our causal-functional brain can discern the sui generis character of qualia. Why do painful qualia cause functional aversion in the brain? Why not the other way around? Why would qualia have functional-causal properties on the brain that match their qualitativity?

We can characterize qualia by how they make us feel. We feel a red color to be hot, bright, intense, beautiful, or similar to orange. But then we could just talk about how we feel. We can posit that qualia are virtual contents of certain functional states we call "sensible experiences". These states operate as if qualia existed, but they do not. This is the meaning of a "virtual" (illusory) object.

[Daniel Coimbra @dan_sci_phil]

Qualia are illusory if the mind can be reduced to something like patterns, functions, operations, structure, causation. Qualia are real if there is 'something extra' in whatever sense.

That is a theoretical way to frame it. One can also frame it semantically in this way. If qualia are illusory then, when we experience red, nothing is red. Instead, all we have a neurofunctional state "as of experiencing red", plus the causal reflectance properties of external materials we call red.

[Daniel Coimbra @dan_sci_phil]

That would be my view, yes. The brain enters a functional state "as if" feeling pain. We might call this state "pain", but the qualia of pain itself does not exist.

Painful experiences are not aversive because they involve painful qualities. Rather, such experience states are labeled as "painful" at all precisely because they operate as functionally aversive to the rest of the brain.

[Daniel Coimbra @dan_sci_phil]

To no one. The cognitive illusion of qualia are the thought patterns as if there was qualia.

My answer raises the question of who is the recipient of a thought's contents. I say a thought's contents are its functional effects on downstream processes: association, memory, behavior, language, further concepts and thoughts, and so on.

[Daniel Coimbra @dan_sci_phil]

@QuantaMagazine @Franfs

[Daniel Coimbra @dan_sci_phil]

@FurkanGozukara @AnthropicAI Honest question: Why think China wouldn't do the same? Their current incentives might dictate open-source, but this can change when China gets AGI models. I don't have an analysis on this.

[Daniel Coimbra @dan_sci_phil]

@discretemath_er @_onionesque If find this one quite legible and fun! About meta-complexity in the theory of computability. That is: Why it is hard or perhaps impossible to prove whether P=NP (for example).
https://t.co/VlFRTVSXgJ