51 Matching Annotations
  1. Last 7 days
    1. Perhaps, he realized, these viruses don’t actually need to unite their segments in the same host cell. “If theory was saying that this is impossible, maybe the viruses just don’t do it,” he says. “And once we had this stupid idea, testing it was very easy.”

      This is different from the theory of evolution or the theory of electromagnetism. It's a smaller things, like an assumption. Evolution, also in biology, is a more encompassing set of ideas. So the theoretical framework has a hierarchy. Perhaps at the top is a Kuhnian paradigm or a Lakatosian research program.

      Does this hierarchy different between sciences, though? Like, how hard is it to take a new assumption and grow it into a fully-fledged theory? Biology is more complex than physics, with more "facts" and forms to understand. Evolution is different from electromagnetism because it doesn't limit as much. EM clearly prescribes what's possible and what isn't, whereas evolution doesn't make the distinction so clearly.

    1. such as scope, simplicity, fruitfulness, accuracy

      Theories can be measured according to multiple metrics. The current default appears to be predictive accuracy, but this lists others, such as scope. If theory A predicts better but narrower and theory B predicts worse (in A's domain) but much more broadly, which is a better theory?

      Others might be related to simplicity and whatnot. For example, if a theory is numerical but not explanatory (such as scaling laws or the results of statistical fitting) this theory might be useful but not satisfying.

    2. Like in evolution, the process does not change toward some fixed goal according to some fixed rules, methods or standards, but rather it changes away from the pressures exerted by anomalies on the reigning theory (Kuhn 1962, 170–173). The process of scientific change is eliminative and permissive rather than instructive.

      This is similar to evolution: not guided, but not random. Does this view contradict the idea of progression?

      It also suggests a complex dynamic system that possess path dependence and environmental interaction.

  2. Mar 2019
    1. The IP metaphor has had a half-century run, producing few, if any, insights along the way.

      The value of an analogy is in the fruit it bears. If no insights have resulted, the analogy is weak.

    2. Worse still, even if we had the ability to take a snapshot of all of the brain’s 86 billion neurons and then to simulate the state of those neurons in a computer, that vast pattern would mean nothing outside the body of the brain that produced it. This is perhaps the most egregious way in which the IP metaphor has distorted our thinking about human functioning.

      Again, this doesn't conflict with a machine-learning or deep-learning or neural-net way of seeing IP.

    3. No ‘copy’ of the story is ever made

      Or, the copy initially made is changed over time since human "memory" is interdependent and interactive with other brain changes, whereas each bit in computer memory is independent of all other bits.

      However, machine learning probably results in interactions between bits as the learning algorithm is exposed to more training data. The values in a deep neural network interact in ways that are not so obvious. So this machine-human analogy might be getting new life with machine learning.

    4. The IP perspective requires the player to formulate an estimate of various initial conditions of the ball’s flight

      I don't see how this is true. The IP perspective depends on algorithms. There are many different algorithms to perform various tasks. Some perform reverse-kinematic calculations, but others conduct simpler, repeated steps. In computer science, this might be dynamic programming, recursive algorithms, or optimization. It seems that the IP metaphor still fits: it's just that those using the metaphor may not have updated their model of IP to be more modern.

    5. just as we now view the hydraulic and mechanical metaphors to be silly

      We certainly view them as silly when compared to our modern perspective, but viewed from the knowledge of the day, it isn't silly. You could also argue that early models of celestial mechanics (that god's pulled the sun across the sky, or something) were silly.

    6. either replaced by another metaphor or, in the end, replaced by actual knowledge

      You can argue that all knowledge is a metaphor of some type. All knowledge is, at best, a simplified map from the intellectual/logical to the empirical. It's just that the metaphor/map might be written in human language, mathematical language, or algorithmic language.

  3. Jan 2019
    1. Bahá’u’lláh states that the knowledge of God is revealed through His Manifestation, Who has an innate awareness of the human condition and the social order

      On behalf of the Universal House of Justice: Baha'u'llah's innate knowledge is of the human reality.

      Is it also of the animal, vegetable, and mineral realities?

    1. This is the meaning of the “Day of Resurrection,” spoken of in all the scriptures, and announced unto all people. Reflect, can a more precious, a mightier, and more glorious day than this be conceived, so that man should willingly forego its grace, and deprive himself of its bounties, which like unto vernal showers are raining from the heaven of mercy upon all mankind?

      I think this meaning is that "Resurrection" is the return of a Manifestation in another human frame. And this is stated to be clearly more glorious than the literal interpretations of past scripture.

      Why is it clearly more glorious?

      1. Everyone has access. And it leads to empowerment.
      2. It allows us to keep science, which is pretty awesome.
      3. It doesn't allow us to just wait for the rapture - see point 1 about empowerment.
      4. It allows us to see all religions as united in spirit.
      5. Related to point 3 and 4, it allows us to unite with non-religious people.
      6. All of this without "doing violence to the facts".
    1. Should a man wish to adorn himself with the ornaments of the earth, to wear its apparels, or partake of the benefits it can bestow, no harm can befall him, if he alloweth nothing whatever to intervene between him and God, for God hath ordained every good thing, whether created in the heavens or in the earth, for such of His servants as truly believe in Him.

      Definition of attachment.

      Also, if your actions are due to addiction but you're consciously trying to counter it (or even, if you don't have help in quitting, just disliking it), perhaps that's different from allowing something to "intervene between him and God".

    2. Know ye that by “the world” is meant your unawareness of Him Who is your Maker, and your absorption in aught else but Him. The “life to come,” on the other hand, signifieth the things that give you a safe approach to God, the All-Glorious, the Incomparable.

      The next world isn't another world. It's a way of being in this world. Using this idea alone, no surviving memory or individuality is required after death.

    3. Say: Teach ye the Cause of God, O people of Bahá, for God hath prescribed unto every one the duty of proclaiming His Message, and regardeth it as the most meritorious of all deeds.

      The discussion in book 6 after this quote explores what a duty is. It presents the idea that, like eating, laws in the Faith to pray or teach are not arbitrary rules, but statements about our nature, and conducive to our growth. This implies that they should also be sources of joy.

    4. Say: Should your conduct, O people, contradict your professions, how think ye, then, to be able to distinguish yourselves from them who, though professing their faith in the Lord their God, have, as soon as He came unto them in the cloud of holiness, refused to acknowledge Him, and repudiated His truth?

      What's the difference between a professed Baha'i who (purposefully?) breaks the laws, and people who do not accept the next Manifestation when They come?

      How does this fit with, for example, alcohol addiction? Is that conduct that is contradicting someone's professions? Maybe not if your professions are "alcohol is bad, we shouldn't do it, but I am addicted and trying to recover", as opposed to "Baha'u'llah says alcohol is bad, but I drink anyway and don't see it as a problem".

    5. Beware lest ye contend with anyone, nay, strive to make him aware of the truth with kindly manner and most convincing exhortation. If your hearer respond, he will have responded to his own behoof, and if not, turn ye away from him, and set your faces towards God’s sacred Court, the seat of resplendent holiness.
    1. A comment at the bottom by Barbara H Partee, another panelist alongside Chomsky:

      I'd like to see inclusion of a version of the interpretation problem that reflects my own work as a working formal semanticist and is not inherently more probabilistic than the formal 'generation task' (which, by the way, has very little in common with the real-world sentence production task, a task that is probably just as probabilistic as the real-world interpretation task).

    2. There is a notion of success ... which I think is novel in the history of science. It interprets success as approximating unanalyzed data.

      This article makes a solid argument for why statistical and probabilistic models are useful, not only for prediction, but also for understanding. Perhaps this is a key point that Noam misses, but the quote narrows the definition to models that approximate "unanalyzed data".

      However, it seems clear from this article that the successes of ML models have gone beyond approximating unanalyzed data.

    3. But O'Reilly realizes that it doesn't matter what his detractors think of his astronomical ignorance, because his supporters think he has gotten exactly to the key issue: why? He doesn't care how the tides work, tell him why they work. Why is the moon at the right distance to provide a gentle tide, and exert a stabilizing effect on earth's axis of rotation, thus protecting life here? Why does gravity work the way it does? Why does anything at all exist rather than not exist? O'Reilly is correct that these questions can only be addressed by mythmaking, religion or philosophy, not by science.

      Scientific insight isn't the same as metaphysical questions, in spite of having the same question word. Asking, "Why do epidemics have a peak?" is not the same as asking "Why does life exist?". Actually, that second question can be interested in two different ways, one metaphysically and one physically. The latter interpretation means that "why" is looking for a material cause. So even simple and approximate models can have generalizing value, such as the Schelling Segregation model. There is difference between models to predict and models to explain, and both have value. As later mentioned in this document, theory and data are two feet and both are needed for each other.

    4. This page discusses different types of models

      • statistical models
      • probabilistic models
      • trained models

      and explores the interaction between prediction and insight.

    5. Chomsky (1991) shows that he is happy with a Mystical answer, although he shifts vocabulary from "soul" to "biological endowment."

      Wasn't one of Chomsky's ideas that humans are uniquely suited to language? The counter-perspective espoused here appears to be that language emerges, and that humans are only distinguished by the magnitude of their capacity for language; other species probably have proto-language, and there is likely a smooth transition from one to the other. In fact, there isn't a "one" nor an "other" in a true qualitative sense.

      So what if we discover something about the human that appears to be required for our language? Does this, then, lead us to knowledge of how human language is qualitatively different from other languages?

      Can probabilistic models account for qualitative differences? If a very low, but not 0, probability is assigned to a given event that we know is impossible from our theory-based view, that doesn't make our probabilistic model useless. "All models are wrong, some are useful." But it seems that it does carry with it an assumption that there are no real categories, that categories change according to the needs, and are only useful in describing things. But the underlying nature of reality is of a continuum.

  4. jasss.soc.surrey.ac.uk jasss.soc.surrey.ac.uk
    1. To Guide Data Collection

      This seems to be, essentially, that models are useful for prediction, but prediction of unknowns in the data instead of prediction of future system dynamics.

    2. Without models, in other words, it is not always clear what data to collect!

      Or how to interpret that data in the light of complex systems.

    3. Plate tectonics surely explains earthquakes, but does not permit us to predict the time and place of their occurrence.

      But how do you tell the value of an explanation? Should it not empower you to some new action or ability? It could be that the explanation is somewhat of a by-product of other prediction-making theories (like how plate tectonics relies on thermodynamics, fluid dynamics, and rock mechanics, which do make predictions).

      It might also make predictions itself, such as that volcanoes not on clear plate boundaries might be somehow different (distribution of occurrence over time, correlation with earthquakes, content of magma, size of eruption...), or that understanding the explanation for lightning allows prediction that a grounded metal pole above the house might protect the house from lightning strikes. This might be a different kind of prediction, though, since it isn't predicting future dynamics. Knowing how epidemics works doesn't necessarily allow prediction of total infected counts or length of infection, but it does allow prediction of minimum vaccination rates to avert outbreaks.

      Nonetheless, a theory as a tool to explain, with very poor predictive ability, can still be useful, though less valuable than one that also makes testable predictions.

      But in general, it seems like data -> theory is the explanation. Theory -> data is the prediction. The strength of the prediction depends on the strength of the theory.

  5. Jul 2018
    1. The empty brain

      To look at what the brain manifests isn't enough. You can find a machine analogy for most of what the brain manifests, but that doesn't mean the analogy is valid, If the brain is actually a computer, then it must not only behave like one, but also response to counter-factuals (experiments) like one. It doesn't. So either the brain is a much more complex computing machine than we ever have, or it is better thought of as something else.

    2. Think how difficult this problem is. To understand even the basics of how the brain maintains the human intellect, we might need to know not just the current state of all 86 billion neurons and their 100 trillion interconnections, not just the varying strengths with which they are connected, and not just the states of more than 1,000 proteins that exist at each connection point, but how the moment-to-moment activity of the brain contributes to the integrity of the system. Add to this the uniqueness of each brain, brought about in part because of the uniqueness of each person’s life history, and Kandel’s prediction starts to sound overly optimistic. (In a recent op-ed in The New York Times, the neuroscientist Kenneth Miller suggested it will take ‘centuries’ just to figure out basic neuronal connectivity.)

      An example of a problem that is both hard to solve and hard to verify. More than NP-hard.

    3. there is no reason to believe that any two of us are changed the same way by the same experience.

      Experiential changes are highly sensitive to the brain's starting conditions - a feature of complex systems.

    4. This might sound complicated, but it is actually incredibly simple, and completely free of computations, representations and algorithms.

      Actually, this sounds exactly like an algorithm, though one we clearly are conscious of. You could build a robot to do this, and in fact, some have. See Perceptual Control Theory by William Powers.

    5. no image of the dollar bill has in any sense been ‘stored’ in Jinny’s brain

      There was research using machine learning to "see" pictures that people imagined. It was able to somewhat decode the measurements of brain activity at various locations outside the head and reconstruct something similar to what was being thought of. True, this isn't a memory, but it could be that the "thought" is "stored" in a pattern of electric potentials, and we just can't decode it. A computer stores information bits without any interaction (ideally). Clearly, human memory isn't so orthogonal. But why couldn't it be some type of encoding into a measurable pattern where, whether it's to save space, improve recall through association, or whatever else?

    6. By the 1500s, automata powered by springs and gears had been devised, eventually inspiring leading thinkers such as René Descartes to assert that humans are complex machines. In the 1600s, the British philosopher Thomas Hobbes suggested that thinking arose from small mechanical motions in the brain. By the 1700s, discoveries about electricity and chemistry led to new theories of human intelligence – again, largely metaphorical in nature. In the mid-1800s, inspired by recent advances in communications, the German physicist Hermann von Helmholtz compared the brain to a telegraph.

      Our most complex machine is always used as an analogy for the brain.

    7. Predictably, just a few years after the dawn of computer technology in the 1940s, the brain was said to operate like a computer, with the role of physical hardware played by the brain itself and our thoughts serving as software.

      This is a very interesting set of correlations, but could it be that our understanding of intelligence is improving precisely because our machines are improving? Our analogies are getting better all the time. Perhaps the current one is "good enough".

    8. knowledge

      We don't develop knowledge ever?

    9. software

      Biological systems of course don't have "software", but isn't this just what gives something it's behaviours and reactions?

    10. rules

      A reflex isn't a rule?

    1. No statistical procedure allows one to somehow see a mundane, taken-for-granted observation in a radically different and new way.

      What makes machine learning unique, perhaps, is that it learns. The question should be about the nature of that learning rather than whether it's possible qualitatively. What about machine learning is not going to match human learning?

    2. But intelligence and rationality are more than just calculation or computation, and have more to do with the human ability to attend to and identify what is most relevant.

      Can this be modelled well with calculation, though? Perhaps the limitation is that we don't know how to code relevance-finding, nor can we train it (in situations that interact with the analog world rather than being explicitly in the simulated).

    3. Kahneman concluded his aforementioned presentation to academics by arguing that computers or robots are better than humans on three essential dimensions: they are better at statistical reasoning and less enamoured with stories; they have higher emotional intelligence; and they exhibit far more wisdom than humans.

      A little over-the-top?

    4. But, while Kahneman calls for large-scale replications of priming studies, the argument here is not that we need more studies or data to verify that people indeed miss blatantly obvious gorillas. Instead, we need better interpretation and better theories.

      More data vs more theories

      Humans are biased by our theories (though not totally). But isn't that the goal of science, to collectively question our assumptions and experiments? We need to attempt to falsify our theories not only by questioning the experiments and repeating them, but also by questioning the theories used to interpret data.

    5. The worry is that the growing preoccupation of many behavioural scientists – across psychology, economics and the cognitive sciences – with blindness and bias causes scientists to look for evidence of human blindness and bias.

      That's not irrational: If someone found gold someplace, I would want to look there, too. A global optimization algorithm will require both trusting your recent results (if you are finding more good stuff here, keep searching here) and distrusting them (to avoid finding a local optimum).

      But this is different because it isn't just about finding the objective optimum, but about interpreting the data itself. If the search strategy is not in data but in data interpretation, though, this analogy might hold. The scientific movement, composed of many scientists, might eventually find the truth, though individuals scientists might get distracted by perceived truth. But what's the measure of truth if not how it manifests in action? ("Judge the tree by the fruit") We thought the geocentric picture was best, but when a small group of people showed that the heliocentric view was much better (in prediction - i.e. a type of scientific action), eventually, the phase-change happened and we left the local optimum.

    6. How you interpret the finding depends on what you are looking for.

      Yes, but not totally. "All things are relative, but some more than others."

    7. After all, many of the most significant scientific discoveries resulted not from reams of data or large amounts of computational power, but from a question or theory.

      There are examples of how data has led to theory as well, Tycho Brahe being a solid one.

    8. ‘omniscience in the observer’
    9. To illustrate, consider Isaac Newton.

      But there are examples of where our theory has led us astray, the heliocentric vision of the universe being an example. If not for that attachment to previous thinking, we might have learned more quickly about the heliocentric truth.

      'Even as He hath revealed: "As oft as an Apostle cometh unto you with that which your souls desire not, ye swell with pride, accusing some of being impostors and slaying others."' - Kitab-i-Iqan

    10. However, computers and algorithms – even the most sophisticated ones – cannot address the fallacy of obviousness. Put differently, they can never know what might be relevant.

      One goal of systems science and modelling, to explore what might be relevant and give us better heuristics.

    11. At the other extreme we have behavioural economics, which focuses on human bias and blindness by pointing out biases or obvious things that humans miss.
    12. So, given the problem of too much evidence – again, think of all the things that are evident in the gorilla clip – humans try to hone in on what might be relevant for answering particular questions. We attend to what might be meaningful and useful

      Consumat, heuristics - actually, this does work with thinking fast and slow. But maybe the divide isn't so clear - a spectrum?

    13. ‘blind to the obvious, and that we also are blind to our blindness’
    14. building on Herbert Simon’s 1950s work on bounded rationality
    15. In other words, there is no neutral observation.

      Similar to the objectivism/subjectivism divide and social construction of truth/positivist scientific method.

      Part of meditative practices is sometimes a focus of the mind, or an absence and an "insistent" self. Perhaps it's simply quieting part of the questioning mind in order to make room for a better side, but maybe it's about removing one's preconceived notions.

    16. The implication (contrary to psychophysics) is that mind-to-world processes drive perception rather than world-to-mind processes.

      Maybe both? Causal feedback.