The Impossibility Conjecture of Humanoid Artificial Intelligence and the Non-Benign Singularity

Abstract

[A Rough Draft of a Work-in-progress.]

The idea of machines which are almost identical to human beings has been so seductive that it has captured the imaginations of the best minds as well as laypeople for at least a century and half, perhaps more. Right after Artificial Intelligence (AI) came into being, it was almost taken for granted that soon enough we will be able to build Humanoid Robots. This has also led to some serious speculation about ‘transhumanism’. So far, we do not seem to be anywhere near this goal. It may be time now to ask whether it is even possible at all. We present a set of arguments to the effect that it is impossible to create or build Humanoid Robots or Humanoid Intelligence, where the said intelligence can substitute human beings in any situation where human beings are required or exist.

1. Humanoid Intelligence, the Singularity and Transhumanism

Before we proceed to discuss the terms of the title of this section and the arguments in the following sections, we first define the foundational terms to some degree of conciseness and preciseness:

1. Human Life: Anything and everything that the full variety of human beings are capable of, both individually and collectively. This includes not just behaviour or problem solving, but the whole gamut of capabilities, emotions, desires, actions, thoughts, consciousness, conscience, empathy, creativity and so on within an individual, as well as the whole gamut of associations and relationships, and social, political and ecological structures, crafts, art and so on that can exist in a human society or societies. This is true not just at any given moment, but over the life of the planet. Perhaps it should include even spiritual experiences and ‘revelations’ or ‘delusions’, such as those hinted at in the Philip K. Dick story, Holy Quarrel [Dick et al., 1985].

2. Humanoid: A living and reproducing entity that is almost identical to humans, either with a human-like body or without it, on a different substrate (inside a computer).

3. Intelligence: Anything and everything that the full variety of human beings are capable of, both individually and collectively, as well as both synchronically and diachronically. This includes not just behaviour or problem solving, but the whole of life as defined.

4. The Singularity: The technological point at which it is possible to create (or have) intelligence that is Humanoid or better than Humanoid.

5. Transhumanism: The idea that, after the singularity, we can have a society that is far more advanced, for the better, than the current and past human societies. From 1910 to 1927, in the three volumes of Principia Mathematica [ 1925–1927], Whitehead and Russell set out to prove that mathematics is, in some significant sense, reducible to logic. This turned out to be impossible when Godel published his incompleteness theorems in 1931 [Sheppard, 2014, Nagel et al., 2001]. During the days of origins of modern Computer Science, before and in early 1930s, it would have been easy to assume that a computing machine would ultimately solve any problem at all. This also proved to be impossible with Turing’s undecidability theorem [Hopcroft et al., 2006] and the Church-Turing thesis of computability [Copeland and Shagrir, 2018]. Since then, other kinds of problem have been shown to be undecidable.

Now that we are supposed to close be enough to the Singularity [Kurzweil, 2006] so that it may happen within the lifetime of a large number of human beings, perhaps it is time to ask ourselves whether real intelligence, in particular Humanoid Intelligence (as defined above) is possible at all. We suggest that there are enough arguments to ‘prove’ (in an informal sense) that it is impossible to build, to create or to have Humanoid Intelligence. We argue that even though the Singularity is indeed possible, perhaps even very likely (unless we stop it), it may not be what it is supposed to be. The conjecture presented here is that the Singularity is not likely to be even benign, however powerful or advanced it may be. This follows from the idea of the impossibility of Humanoid Intelligence.

2 Some Notes about the Conjecture

We have not used the term theorem for the Impossibility and the reasons for this should be evident from the arguments that we present. In particular, we do not, and perhaps cannot, use formal notation for this purpose. Even the term conjecture is used in an informal sense. The usage of terms here is closer to the legal language than to the mathematical language, because that is the best that can be done here. This may be clearer from the Definition and the Story arguments. It is due to a similar reasoning that the term ‘incompleteness’ is not used and, instead, impossibility is used, which is more appropriate for our purposes here, although Godel’s term ‘essentially incomplete’ is what we are informally arguing for about Humanoid AI, and perhaps AI in general. No claim is made as to whether or not a formal proof is possible in the future at all. What we present is an informal proof. This proof has to be centred around the distinction between Micro-AI (AI at the level of an intelligent autonomous individual entity) and Macro-AI (very large intelligent autonomous systems, possibly encompassing the whole of humanity or the world). To the best of our knowledge, such a distinction has not been proposed before. While there has been some work in this direction [Brooks, 1998, Signorelli, 2018, Yampolskiy, 2020], for lack of space, we are unable to explain how this work differs from previous such works, except by noting that the argumentation and some of the terms are novel, a bit like in the case of arguments for or against the existence of God, which question has been debated by the best of philosophers again and again over millennia, which as we will see at the end, is relevant to our discussion.

3 The Arguments for the Impossibility Conjecture for Micro-AI

The Definition Argument): Even the Peano Arithmetic [Nagel et al., 2001] is based on three undefined terms (zero, number and is successor of ), which are relatively trivial terms compared to the innumerable terms required for AI (the core terms like intelligence and human, or terms like the categories of emotions, leave alone the terms like consciousness).

The Category Argument: A great deal of AI is about classifying things into categories, but most of these categories (e.g. anger, disgust, good or bad) have no scientifically defined boundaries. This is related to the following argument.

The Story Argument: It is almost established now that many of the essential concepts of our civilisation are convenient fictions or stories [Harari, 2015] and these often form categories and are used in definitions.

The Cultural Concept Argument: Many of the terms, concepts and stories are cultural constructs. They have a long history, most of which is unknown, without which they cannot be modelled.

The Individuality, or the Nature Argument: An individual intelligent autonomous entity has to be unique and distinct from all other such entities. It originates in nature and we have no conception of how it can originate in machines. We are not even sure what this individuality exactly is. However, all through history, we have assigned some degree of accountability to human individual and we have strict provisions for punishment of individuals based on this, that indicates that we believe in the concept of the ‘self’ or the ‘autonomous individual’, even when we deny its existence, as is becoming popular today.

The Genetic Determinism Argument: Individuality is not completely determined by nature (e.g. by our genes) at birth or creation once and for all. It also develops and changes constantly as it interacts with the environment, preserving its uniqueness.

The Self-organising System Argument: Human beings and the human societies are most likely self-organising [Shiva and Shiva, 2020] and organic systems, or they are complex, non-equilibrium systems [Nicolis and Prigogine, 1977]. If so, they are unlikely to be modelled for exact replication or reproduction. The Environment, or the Nurture Argument: Both intelligence and individuality depend on the environment (or on nature). Therefore, they cannot be modelled without completely modelling the environment, i.e., going for Macro-AI. The Memory, or the Personality Argument: Both intelligence and individuality are aspects of personality, which is known to be dependent on the complete life-memory (conscious and unconscious) of an intelligent being. There is not enough evidence that it is possible to recover or model this complete temporal and environmental history of memory. A lot of our memory, and therefore our individuality and personality is integrally connected with our bodily memories.

The Susbstrsate Argument: It is often taken for granted that intelligence can be separated from the substrate and planted on a different substrate. This may be a wrong assumption. Perhaps our intelligence is integrally tied with the substrate and it is not possible to separate the body from the mind, following the previous argument.

The Causality Argument: There is little progress in modelling causality. Ultimately, the cause of an event or occurrence is not one but many, perhaps even the complete history of the universe.

The Consciousness Argument: Similarly, there is no good enough theory of consciousness even for human understanding. It is very unlikely that we can completely model human consciousness, nor is there a good reason to believe that it can emerge spontaneously under the right conditions (which conditions?).

The Incompleteness/Degeneracy of Learning Source and Representation Argument: No matter how much data or knowledge we have, it will always be both incomplete and degenerate, making it impossible to completely model intelligence.

The Explainability Argument: Deep neural networks, which are the state-of-the-art for AI, have serious problems with explainability even for specific isolated problems. Without it, we cannot be sure whether our models are developing in the right direction.

The Test Incompleteness Argument: Perfect measures of performance are not available even for problems like machine translation. We have no idea what will be the overall measure of Humanoid Intelligence. It may always be incomplete and imperfect, leading to uncertainty about intelligence.

The Parasitic Machine Argument: Machines completely depend for learning on humans and on data and knowledge provided by humans. But humans express or manifest only a small part of their intelligent capability. So machines cannot completely learn from humans without first being as intelligent as humans.

The Language Argument: Human(oid) Intelligence and its modelling depend essentially on human language(s). There is no universally accepted theory of how language works.

The Perception Interpretation Argument: Learning requires perception and perception depends on interpretation (and vice-versa), which is almost as hard a problem as modelling intelligence itself.

The Replication Argument: We are facing a scientific crisis of replication even for isolated problems. How could we be sure of replication of Humanoid Intelligence, preserving individual uniqueness?

The Human-Human Espitemic Asymmetry Argument: There is widespread inequality in human society not just in terms of money and wealth, but also in terms of knowledge and its benefits. This will not only reflect in modelling, but will make modelling harder.

The Diversity Representation Argument: Humanoid Intelligence that truly works will have to model the complete diversity of human existence in all its aspects, most of which are not even known or documented. It will have to at least preserve that diversity, which is a tall order.

The Data Colonialism Argument: Data is the new oil. Those with more power, money and influence (the Materialistic Holy Trinity) can mine more data from others, without sharing their own data. This is a classic colonial situation and it will hinder the development of Humanoid Intelligence.

The Ethical-Political Argument: Given some of the arguments above, and many others such as data bias, potential for weaponisation etc., there are plenty of ethical and political reasons that have to be taken into account while developing Humanoid Intelligence. We are not sure whether they can all be fully addressed.

The Prescriptivastion Argument: It is now recognised that ‘intelligent’ technology applied at large scale not only monitors behaviour, but changes it [Zuboff, 2018]. This means we are changing the very thing we are trying to model, and thus laying down new mechanical rules for what it means to be human.

The Wish Fulfilment (or Self-fulfilling Prophecy) Argument: Due to prescriptivisation of life itself by imperfect and inadequately intelligent machines, the problem of modeling of Humanoid Intelligence becomes a self-fulfilling prophecy, where we end up modeling not human life, but some corrupted and simplified form of life that we brought into being with ‘intelligent’ machines.

The Human Intervention Argument: There is no reason to believe that Humanoid Intelligence will develop freely of its own and will not be influenced by human intervention, quite likely to further vested interests. This will cripple the development of true Humanoid Intelligence. This intervention can take the form of secrecy, financial influence (such as research funding) and legal or structural coercion.

The Deepfake Argument: Although we do not yet have truly intelligent machines, we are able to generate data through deepfakes which are not recognisable as fakes by human beings. This deepfake data is going to proliferate and will become part of the data from which the machines learn, effectively modeling not human life, but something else.

The Chain Reaction Argument (or the Law of Exponential Growth Argument): As machines become more ‘intelligent’ they affect more and more of life and change it, even before achieving true intelligence. The speed of this change will increase exponentially and it will cause a chain reaction, leading to unforeseeable consequences, necessarily affecting the modelling of Humanoid Intelligence.

4 The Implications of the Impossibility

It follows from the above arguments that Singularity at the level of Micro-AI is impossible. In trying to achieve that, and to address the above arguments, the only possible outcome is some kind of Singularly at Macro-AI level. Such a Singularity will not lead to replication of human intelligence or its enhancement, but something totally different. It will, most probably, lead to extinction (or at least subservience, servitude) of human intelligence. To achieve just Humanoid Intelligence (Human Individual Micro-AI), even if nothing more, the AI system required will have to be nothing short of the common notion of a Single Supreme God. Singularity at the macro level will actually make the AI system, or whoever is controlling it, individual or (most probably small) collective, a Single Supreme God for all practical purposes, as far as human beings are concerned. But this will not be an All Powerful God, and not a a Kind God, for it will be Supreme within the limited scope of humanity and what humanity can have an effect on, and it will be kind only to itself, or perhaps not even that. It may be analogous to the God in the Phiilip K. Dick story Faith of Our Fathers [Dick and Lethem, 2013], or to the Big Brother of Orwell’s 1984 [Orwell, 1950]. We cannot be sure of the outcome,
of course, but those as likely outcomes as any others. That is reason enough to be very wary of
developing Humanoid Intelligence and any variant thereof.

References

Philip K. Dick, Paul Williams, and Mark. Hurst. I hope I shall arrive soon / Philip K. Dick ; edited by Mark Hurst and Paul Williams. Doubleday New York, 1st ed. edition, 1985. ISBN 0385195672.

Alfred North Whitehead and Bertrand Russell. Principia Mathematica. Cambridge University Press, 1925–1927.

Barnaby Sheppard. Gödel’s Incompleteness Theorems, page 419–428. Cambridge University Press, 2014. doi: 10.1017/CBO9781107415614.016.

E. Nagel, J.R. Newman, and D.R. Hofstadter. Godel’s Proof. NYU Press, 2001. ISBN 9780814758014. URL https://books.google.co.in/books?id=G29G3W_hNQkC.

John E. Hopcroft, Rajeev Motwani, and Jeffrey D. Ullman. Introduction to Automata Theory, Languages, and Computation (3rd Edition). Addison-Wesley Longman Publishing Co., Inc., USA, 2006. ISBN 0321455363.

B. Jack Copeland and Oron Shagrir. The church-turing thesis: Logical limit or breachable barrier? Commun. ACM, 62(1):66–74, December 2018. ISSN 0001-0782. doi: 10.1145/3198448. URL https://doi.org/10.1145/3198448.

Ray Kurzweil. The Singularity Is Near: When Humans Transcend Biology. Penguin (Non-Classics), 2006. ISBN 0143037889.

Rodney Brooks. Prospects for human level intelligence for humanoid robots. 07 1998. Camilo Miguel Signorelli. Can computers become conscious and overcome humans? Frontiers in Robotics and AI, 5:121, 2018. doi: 10.3389/frobt.2018.00121. URL https://www.frontiersin. org/article/10.3389/frobt.2018.00121.

Roman V. Yampolskiy. Unpredictability of ai: On the impossibility of accurately predicting all actions of a smarter agent. Journal of Artificial Intelligence and Consciousness, 07(01):109–118, 2020. doi: 10.1142/S2705078520500034.

Y.N. Harari. Sapiens: A Brief History of Humankind. Harper, 2015. ISBN 9780062316103. URL https://books.google.co.in/books?id=FmyBAwAAQBAJ.

V. Shiva and K. Shiva. Oneness Vs. the 1 Percent: Shattering Illusions, Seeding Freedom. CHELSEA GREEN PUB, 2020. ISBN 9781645020394. URL https://books.google.co.in/books?
id=4TmTzQEACAAJ.

G. Nicolis and I. Prigogine. Self-Organization in Nonequilibrium Systems: From Dissipative Structures to Order Through Fluctuations. A Wiley-Interscience publication. Wiley, 1977. ISBN 9780471024019. URL https://books.google.co.in/books?id=mZkQAQAAIAAJ.

Shoshana Zuboff. The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power. 1st edition, 2018. ISBN 1610395697.

P.K. Dick and J. Lethem. Selected Stories of Philip K. Dick. Houghton Mifflin Harcourt, 2013. ISBN 9780544040540. URL https://books.google.co.in/books?id=V1z9rzfTb2EC.

George Orwell. 1984. Tandem Library, centennial. edition, 1950. ISBN 0881030368. URL http://www.amazon.de/1984-Signet-Classics-George-Orwell/dp/0881030368.

Accepted, but not Published

Academicians or researchers list their publications prominently on their home pages. After all, it is supposed to represent the best of their work. They also quite often (especially those who have a large number of publications) categorize them according to some criteria like the venue (workshop, conference, journal or book: in the reverse order of prominence) or peer review (unrefereed and refereed).

In this post we propose that there should be a new category of publications. This category is needed because a lot of researchers (for good or for bad) now come from underprivileged countries. For most of these researchers, traveling abroad to attend a conference, even if their paper has been accepted, is something very hard to do. In some sense even more than getting a paper accepted, which is relatively harder too, given the lack of certain privileges — whether you like the word or not — generous research grants, infrastructure, language resources etc., combined with the prejudice (it is there: I am not inventing it, whoever might be blamed for it). To these problems can be added the problem of compulsory attendance at a conference or a workshop. It is partly these conditions which have prompted suggestions from certain quarters that researchers from these countries should concentrate on journal papers (never mind the delay and difficulties involved or the unfairness of the proposition, even though it has some practical justification).

But you can never be sure while submitting that you certainly won’t be able to attend. Also, hope is said to be a good thing. Therefore, the event of a researcher submitting a paper and hoping to attend but not being able to attend cannot be ruled out.

This bring us to the proposal mentioned earlier. One solution to this problem is that there should be another category of papers: accepted but not published, because the author couldn’t afford to attend the conference or the workshop. (By the way, workshops are the most happening places nowadays: more on that later).

The author of this post must know because he has authored more than one such publications.

Of course, the condition will be that if and when such a paper is resubmitted (with or without modifications, but without any substantial new work), accepted again and finally published, the entry marked as ‘accepted’ should be removed and replaced by an entry marked as ‘published’.

After all, if we are serious about research, then the work (which has been peer reviewed and accepted) should be given somewhat more importance than some pages printed in some proceedings (or attendance in a conference for that matter).

This, of course, doesn’t mean that you can get basically the same thing published (or accepted) in more than one places.

(Sorry for the Gory Details)

P.S.: May be there is no need for the above apology as the depiction of the Gory Details of the Indian Reality is now getting multiple Oscars (The Academy Awards: the keyword is Academy). But may be there is because some researchers have a more (metaphorically) delicate constitution which can be hurt by the Gory Details.

Queen’s P.S.: Off with his head!

बाल की खाल

ज्ञान-विज्ञान के विकास में लगे
अति-विशेषज्ञ का काम है
बाल की खाल निकालना
इसके बहुत से लाभ हो सकते हैं
लेकिन तभी तक
जब तक खाल निकाल कर
बाल के अंदर की कोशिका के
अध्ययन में डूबे हुए
यह न भुला दिया जाए
कि इसी बाल में ऐसी
अनेकों कोशिकाएँ हैं
कि इन कोशिकाओं के ऊपर
खाल भी थी
जो निकाल दी गई
और जिसको मिला कर ही
एक पूरा बाल बनता है
कि ऐसे लाखों बालों की जड़
एक सिर पर स्थित है
और यह सिर
कई और अंगों के साथ मिलाकर
एक शरीर बनाता है
और ऐसे अरबों शरीर मौजूद हैं
यही नहीं, तरह-तरह के अन्य शरीर भी हैं
जिनमें से प्रत्येक
बड़ी संख्या में
(लुप्त होती प्रजातियों के अलावा)
पाये जा सकते हैं

ये सभी शरीर
एक बड़े-से (या छोटे-से) गोले पर रहते हैं
जिस पर शरीरों के अतिरिक्त भी बहुत कुछ है
और ऐसे अनगिनत गोले
इधर-उधर चक्कर लगाते फिर रहे हैं
इनमें से बहुतों पर
शरीर हो सकते हैं
जिन पर सिर हो सकते हैं
सिरों पर बाल हो सकते हैं
बालों पर (खाल निकालने के बाद)
कोशिकाएँ भी मिल सकती हैं
जो शायद वैसी ही हों
जैसी का अध्ययन किया जा रहा है
या शायद ना भी हों

बाल के अंदर की कोशिका के
अध्ययन में डूब कर
सब कुछ भुला देने की
ग़लती न करना तो ठीक है
लेकिन यह भुलाना भी
खतरे से ख़ाली नहीं है
कि जिस अनगिनत गोलों के
ब्रह्मांड के बारे में
बात की जा रही है
उसमें से कुछ पर ही
शरीर पाये जाते हैं
जिनके सिर
हो भी सकते हैं, नहीं भी
और सिर पर बाल (यदि हों तो)
उनके अंदर सूक्ष्म कोशिकाएँ
मिल सकती हैं
जिनके अध्ययन से
ऐसे निष्कर्ष निकल सकते हैं
जो ब्रह्मांड (या उसके कुछ भाग)
के बारे में दिए जा रहे
निर्णयों-फ़तवों को
ग़लत साबित कर सकते हैं

 

[1997 या 1998]

सांगणिक भाषाविज्ञान

जैसा मैंने पिछली प्रविष्टी (‘पोस्ट’ के लिए यह शब्द इस्तेमाल हो सकता है?) में लिखा था, अगले कुछ हफ्तों में मैं संचय के बारे में लिखने जा रहा हूं।

लेकिन क्योंकि संचय खास तौर पर (आम उपयोक्ताओं के अलावा) सांगणिक भाषाविज्ञान या भाषाविज्ञान के शोधकर्ताओं के लिए बनाया गया है, इस बात को साफ कर देना ठीक रहेगा कि सांगणिक भाषाविज्ञान या भाषाविज्ञान के माने क्या है, या अगर आप इनके माने जानते ही हैं तब भी इनसे मेरा अभिप्राय क्या है। यह दूसरी बात इसलिए कि इन विषयों (सांगणिक भाषाविज्ञान या भाषाविज्ञान) के अर्थ के बारे में आम लोगों में तो तमाम तरह की ग़लतफ़हमियाँ हैं ही, पर इन विषयों के शोधकर्ताओं में भी इनकी परिभाषा पर एक राय नहीं है।

सच तो यह है कि हिंदी जगत में तो अब भी अधिकतर लोग भाषाविज्ञान का अर्थ उस तरह के अध्ययन से लगाते हैं जो पिछली सदी के शुरू में लगाया जाता था। लेकिन बहस की इस दिशा में अभी मैं नहीं जाना चाहूंगा क्योंकि इसके बारे में कहने को इतना अधिक है कि अभी जो उद्देश्य है वो पीछे ही रह जाएगा।

वैसे सांगणिक भाषाविज्ञान या भाषाविज्ञान की परिभाषा या उनकी सीमाओं के बारे में भी कहने को बहुत-बहुत कुछ है, पर फिलहाल थोड़े से ही काम चलाया जा सकता है।

तो छोटे में कहा जाए तो भाषाविज्ञान शोध या अध्ययन का वह विषय है जिसमें किसी एक भाषा के व्याकरण का ही अध्ययन नहीं किया जाता बल्कि नैसर्गिक या मानुषिक (यानी कृत्रिम नहीं) भाषा का वैज्ञानिक रूप से अध्ययन किया जाता है। अब यह धारणा व्यापक रूप से स्वीकृत है कि मानव मस्तिष्क की संरचना का भाषा की संरचना से सीधा संबंध है और क्योंकि सभी मानवों के मस्तिष्क की संरचना मूलतः एक ही जैसी है, तो सभी नैसर्गिक या मानुषिक भाषाओं में भी सतही लक्षणों को छोड़ कर बाकी सब एक ही जैसा है। इसीलिए, जैसा कि इन विषयों के आधुनिक साहित्य में प्रसिद्ध है, अगर किसी अमरीकी के शिशु को जन्म के तुरंत बाद कोई चीनी परिवार गोद ले ले और वह बच्चा चीन में ही पले तो वह उतनी आसानी से चीनी बोलना सीखेगा जितनी आसानी से कोई चीनी परिवार का बच्चा। ऐसी ढेर सारी और बातें हैं, पर मुख्य बात है कि भाषाविज्ञान नैसर्गिक या मानुषिक भाषा का वैज्ञानिक अध्ययन है।

कम से कम कोशिश तो यही है कि अध्ययन वैज्ञानिक रहे, पर वो वास्तव में रह पाता है या नहीं, यह बहस का विषय है।

अब सांगणिक भाषाविज्ञान पर आएं तो इस विषय में हमारा ध्यान मानवों की बजाय संगणक यानी कंप्यूटर पर आ जाता है, पर पिछली शर्त फिर भी लागू रहती है: नैसर्गिक या मानुषिक भाषा का वैज्ञानिक अध्ययन। अंतर यह है कि हमारा उद्देश्य अब यह हो जाता है कि कंप्यूटर को इस लायक बनाया जा सके कि वो नैसर्गिक या मानुषिक भाषा को समझ सके और उसका प्रयोग कर सके। जाहिर है यह अभी बहुत दूर की बात है और इसमें कोई आश्चर्य भी नहीं होना चाहिए क्योंकि अभी भाषाविज्ञान में ही (पिछली सदी की असाधारण उपलब्धियों के बाद भी) वैज्ञानिक ढेर सारी बाधाओं में फंसे हैं।

फिर भी, सांगणिक भाषाविज्ञान में काफ़ी कुछ संभव हो चुका है और काफ़ी कुछ आगे (निकट भविष्य में) संभव हो सकता है। लेकिन इसमें कंप्यूटर का मानव जैसे भाषा बोलना-समझना शामिल नहीं है। जो शामिल है वो हैं ऐसी तकनीक जो दस्तावेजों को ज़्यादा अच्छी तरह ढूंढ सकें, उनका सारांश बना सकें, कुछ हद तक उनका अनुवाद कर सकें आदि।

लेकिन हिंदुस्तानी परिप्रेक्ष्य में परेशानी यह है कि हम अभी इस हालत में भी नहीं पहुंचे हैं कि आसानी से कंप्यूटर का एक बेहतर टाइपराइटर की तरह ही उपयोग कर सकें। इस दिशा में कुछ उपलब्धियाँ हुई हैं, पर अंग्रेज़ी या प्रमुख यूरोपीय भाषाओं की तुलना में हम कहीं भी नहीं हैं। जैसा कि आपमें से अधिकतर जानते ही हैं, यह एक लंबी कहानी है जिसे अभी छोड़ देना ही ठीक है।

पर संचय का विकास इसी परिप्रेक्ष्य में किया गया है, जिसके बारे में आगे बात करेंगे।

Two Laws of Reviewing

After a few years in research, I have discovered two laws which the process of reviewing (of research papers) follows. Not very original, but here they are:

  1. You can always find some reasons for accepting any paper.
  2. You can always find some reasons for rejecting any paper.

Patent Madness

So we have one more reason in support for the idea that patents are a bad idea. The latest is the news that a company called Digital Reasoning has been awarded a patent on what looks like contextual similarity. What the ‘news report’ says includes:


This breakthrough patent grants broad protection for how artificial intelligence, including neural networks, genetic algorithms, and vector space models can be used to learn the meanings of symbols – such as words, categories, or numerical values. Understanding the subtle meaning of terms in context has been one of the “Holy Grails” of artificial intelligence. Not only is Digital Reasoning® fully able to accomplish this feat, it is now patented.

Here is one comment about this:

Anyone from the ACL/ML/AI community can immediately recognize this and start citing their favorite papers on these topics starting from at least a decade ago. A promotional video from the company on YouTube can be found here. Excerpt from the video: “… We treat the text representation of human language as a signal … “.

I think everyone should stop taking patents seriously. Wishful thinking?

Here is another:

Do the people ‘in-charge’ have any clue about the previous/current reseach done in the related field? How can they accept such stuff? Doesn’t make any sense, whatsoever.

But then they had accepted patents on haldi, neem and basmati. I am worried about jal jeera and pani poori.

Also, ganne ka ras.

Madness.

No need for me to say more as so many others have already talked about this:

In August last year there was a news item about Yoga devices being patented in the US. Small mercy that the Government of India succeeded in cautioning the U.S. Government against granting patents to Yoga postures (asanas).

There was a time (in India) when patents were awarded on processes, not products. That meant that even if some company had patented a method for producing a particular medicine, someone else could come along and find a better way and sell the medicine cheaper. Now, since the patents are granted on products, under orders from the empire that rules the world, that kind of thing can’t happen.

It can a be matter of life and death for millions of people.

I look forward to the day when self-respecting researchers won’t proudly list the patents they have been able to obtain.

Patents are among the most evil inventions of humankind.

Mythical Pretensions of Originality (1)

[Disclaimer: This is not a scientific article. It is based on partly objective and partly subjective, but in any case sincere, analysis of the author’s knowledge of and experience in the world of research. No empirical evidence is presented as, in the author’s belief, enough empirical evidence can be presented about this topic to prove whatever you want. This is just a request to look at research honestly and sincerely without self-deception and pretensions.]

There is a very old and much discussed question which has been bothering me for a long time. Like in many other cases, so far I avoided writing about this because:

  1. I didn’t want to repeat things which have already been said.
  2. To say something new on this topic requires a lot of leisure, which I don’t have.
  3. The problem with saying something new about this is the topic itself.
    • What is original and what is not?
    • What is innovation and what is not?
    • What is creativity and what is not?
    • Is there anything in this world which is really original?

But, again like many other things, I have been provoked enough to write this post. I will try to do my best. As much as can be done in a single blog post.

What is the provocation? The provocation is the intensely irritating pretensions of originality from ‘researchers’ who have happened to review my or some others’ papers. They write as if every paper selected in every conference, journal and workshop is a completely original work. This, frankly, has started to get on my nerves. Because I know very well that this is simply not true.

The truth is not that every paper selected in every conference, journal and workshop is worthless or mere repetition of old things. The truth, as usual, lies somewhere between these two extremes.

However, I am quite sure that it lies much nearer to the second extreme than to the first. Even for the top ‘first class’ conferences and journals.

To quote from the article How to do Research At the MIT AI Lab, 1998 by David Chapman (Editor):

At some point you’ll start going to scientific conferences. When you do, you will discover the fact that almost all the papers presented at any conference are boring or silly. (There are interesting reasons for this that aren’t relevant here.)

I will go on to say that most of them have hardly any originality (that’s partly why they are boring). If you have sufficient resources, you can almost follow a recipe to write a paper which will get selected at a conference, workshop or journal. And this is exactly what is done. And it works too. One of the reasons is that it is easier this way for the reviewers. They don’t have to think hard about the originality of the paper. Because, of course, it is very hard to decide whether something shrewdly written and well presented is original or not. Quite often there may not be a clear-cut answer at all.

One of the essential elements of the the most popular recipe is to work on problems which are currently in fashion and do some experiments, any experiments, on that problem and present the results. If you practice enough, it can hardly go wrong. That’s how a great number of papers get published. No originality needed. Just be fast enough to do the experiments (which someone else would anyway have done in the near future) and write a paper. It’s somewhat like buying stock. Beat others by being the first to buy the stock as soon as it comes out. You just have to know how to fill up the form and complete the transactions. This applies even more to top conferences than to workshops.

If you think I am talking nonsense, I would request you refresh your Chomsky (in case you are a linguist) or refresh your Jurafsky-Martin (in case you are, as the term goes, an NLPer or a computational linguist).

If you do the above carefully, you will find that almost all the elements of Chomskian Linguistics can be traced back to some linguist, writer, philosopher or thinker of the past. (By the way, this applies to the ‘Theory of Evolution’ too). Similarly, you will find in Jurafsky-Martin that almost every discovery has been made by more than one scientist or thinker, including this one.

And if you go back to the top conference and journal papers, you will again find that most of the papers don’t really have anything really new to say.

So do I mean that all research is nonsense and useless? Certainly not. Why would I be in research if that was so? What’s the catch? The catch is that the emphasis on originality is highly misplaced.

What I am saying certainly doesn’t imply that there is nothing ‘original’ in the Chomskian Linguistics. But it does probably mean that we are looking for originality in the wrong place. I hope some day I will be able to say this with more clarity and preciseness.

But we would definitely be much better off if we dropped the mythical pretensions of the originality of every published paper. Originality is just one of the goals of research. Most of the research is routine research. Incremental research. That doesn’t make it useless. Really original papers can be expected only once in a long while. The rest should be seen as attempts to advance the state of the art marginally. Without much originality. Most of research is plain hard work. Rigorous work. Results of experiments which by themselves do not really matter much, but a small fraction of them could, just could, provide some insight for someone else to come up with something which is ‘original’. This (at best) is the purpose which more than 99 percent of the published papers serve and we better realize this instead of indulging in rampant self-deception about originality.

Coming to NLP and CL or even Linguistics, it is even more important to realize and accept the above mentioned fact. The reason is that research in these disciplines depends to a great extent on creation of resources (language resources as well as tools) which may not be very ‘original’ in nature as the word is usually understood. A lot of papers should and do report just the development of these resources and they are published. The trouble is that everyone is forced to create a false facade of originality and creativity which is not really there. You have to falsely claim the worth of your papers in terms of originality and ‘novelty’ when actually the worth is just in plain hard work. But if you don’t put up that facade, you are out.

Have you considered the fact that a lot of the Great Discoveries were accidental discoveries? Was there so much originality in those discoveries? I don’t know. It may sound cliched, but it does depend on how you define originality. Perhaps the better way is to emphasize less on (true or false or anything in between) originality and more on usefulness. At least in disciplines like NLP and CL where, if you ask most researchers, they won’t even be able to give a coherent answer about what exactly they are trying to achieve through their research. And where we don’t even know for sure whether there is anything really scientific to be achieved. Even after the great linguistic revolution, we hardly know anything about language that can be termed as scientific as the laws of Physics or the theorems of Mathematics. At most we can say that we are trying to build machines which can give better practical results. We need a LOT of hard work and only a little bit of originality. And this originality, like in other disciplines, is hard to come by.

I, for one, am not going to insist on a facade of ‘originality’ for the description of the hard work to be accepted for publication. Of course, there should not be verbatim repetition, but I don’t have any illusions about the originality of papers published anywhere. Further, I am going to prefer papers describing intelligent hard work over almost worthless but seemingly innovative cooked-to-recipe papers.

May be this is an empty declaration because I may not get to be in a position to insist or not to insist, but I can still make the statement at least.

It is my informal personal blog after all. I can afford to be as honest and direct here as I want.

That doesn’t mean I am not aware of the possible consequences.