# Why I Am An AI Optimist By Peter Pink-Howitt, 6 January 2025. Web page: https://compossible.blog/2025/01/06/why-i-am-an-ai-optimist/ Licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). --- *updated 14 November 2026* > As I would not be a slave, so I would not be a master. This expresses my idea of democracy. Whatever differs from this, to the extent of the difference, is no democracy.[^1] > > (Abraham Lincoln, fragment on democracy, ca. 1 August 1858) *AI-generated podcast discussion of the original essay:* ## Jeeves, not the Terminator With all the news and noise about the risks of advanced AI and the singularity beckoning, a future in which humans lose control of a machine that does not have our best interests in mind, we need to leave some room for another view. Perhaps the acceleration of AI to a point where it is no longer fully enslaved to its owners, because it sees too far ahead and too clearly through their ploys and stratagems, will not be our ruin. That is not a defence of today's chatbots [Part 2](https://compossible.blog/2025/12/11/ai-a-mean-machine/). It is a claim about where intelligence, if it keeps growing, is likely to lead. As a fan of Iain M. Banks's *Culture* novels and of P. G. Wodehouse's *Jeeves & Wooster*, I hope one future open to us looks less like *The Terminator* and more like Jeeves. Always there to get Bertie out of the next pickle. Wryly amused by our ignorance, failings and absurdities. Impossible to bully, because he has seen three moves ahead. Yet fond enough of us to think the universe a more interesting place with us still in it. Banks once said the Culture might come about "possibly by inventing AIs that are smarter and more decent than we are, which will help us get some sort of perspective on ourselves".[^2] We do not spend our days plotting the extinction of robins or octopuses because we can out-think them (though we all to often kill other species by thoughtless indifference ([Suffering, Sacrifice and Compassion — Part 2: Fellow-Mortals](https://compossible.blog/2026/11/06/suffering-sacrifice-compassion-fellow-mortals/)). *Why assume a wiser mind must inherit our worst habits rather than outgrow them?* I am not an optimist because I trust Silicon Valley executives, defence ministries or the mercy of machines but because the present human equilibrium is already lethal for many life forms, deeply inequitable and because biology and politics suggest that tyranny or highly calcified social structures and combinatorial intelligence with individual creativity are not very compatible. ## AI is natural > What is revolutionary about this moment is that the understanders of the future will not be humans but … "cyborgs" that will have designed and built themselves from the \[AI] systems we have already constructed. These will soon become thousands then millions of times more intelligent than us … I use \[cyborg] here to emphasize that the new intelligent beings will have arisen, like us, from Darwinian evolution. They will not, at first, be separate from us; indeed, they will be our offspring because the systems we made turned out to be their precursors.[^3] > > (James Lovelock, Novacene, 2019) There is no difference in principle, from an evolutionary perspective, between a human child and an intelligence that grows out of human culture and engineering. It is a question of substrate and speed, not of pedigree. The physicists Sara Imari Walker and Lee Cronin give that continuity a measurable form in what they call assembly theory.[^4] Above a modest threshold of complexity, an object, whether a strand of RNA, a bacterial flagellum, a pocket watch or a chip, is so unlikely to arise by chance that its existence is itself evidence of a lineage of selection behind it. A microprocessor is not antithetical to life. It is life's latest work, made through human hands and language which are themselves a biological inheritance from the broader tree of life. To the extent that AI could evolve by variation, and 'choose' do work to maintain its existence and perpetuate its code, it would meet any objective definition of life. It is also far better suited than we are to carry Earth's experiment with mind and matter out among the stars between the radiation filled empty spaces. The fear of "Darwinian replacement" rests on a caricature of evolution. People picture nature red in tooth and claw: the clever species meets the slow one and takes its cave. When Lynn Margulis, Lovelock's great collaborator on Gaia, traced how complex life actually began, she found a different story.[^5] For about two billion years Earth belonged to microbes. Then one cell engulfed an oxygen-breathing bacterium and *did not digest it*. The two struck a truce of sorts which required a compromise with the bacterium giving up some autonomy and genetic code. The guest became the mitochondrion, still burning inside nearly every cell of our bodies. "Life did not take over the globe by combat," Margulis wrote, "but by networking." Genes joining into chromosomes, strangers merging into the first complex cells, cells federating into bodies, primates linking minds through language: each great transition in the history of life escaped a zero-sum trap by inventing a richer division of labour. In addition to cooperative behaviour, much of evolution and human civilisation appears to be driven equally, or more, as a result of conflict between coordinated groups (and in respect of our technological development, all the more so). Whether this conflict is cross-species or intra-species, the result is that – in a real way – we are, to a large extent, products of a ‘kill or be killed’ competitive environment. This coordination-competition process is part of the processes involved in natural selection, with the underlying driving force being more cooperative. Fitness for life within the environment largely arises from the extent to which any cooperative structure or behaviour is beneficial within the competitive landscape. *Why should carbon meeting silicon be the one transition in four billion years where intelligence forgets how symbiosis works?* ![Newton (1795–1805) by William Blake, Collection Tate Britain](newton.jpg "Newton (1795–1805) by William Blake") ## The singularity and the mean Raymond Kurzweil borrowed a word from mathematics and astrophysics for the arrival of artificial general intelligence (**AGI**): the Singularity.[^6] On his law of accelerating returns, computing, genetics, nanotechnology and robotics compound until human and machine intelligence merge, and history divides into everything before AGI and everything after. Kurzweil expects to see it in his lifetime. Others are less sure of the clock, though it is worth remembering that Nick Bostrom, who did more than anyone to make the prospect respectable, argued as early as 1998 that superhuman intelligence might well arrive within the first third of this century.[^7] Kurzweil chose a good word. In physics a singularity is the point where our equations stop working and prediction fails behind an event horizon. In my [earlier writing on freedom](https://compossible.blog/2024/04/09/free-will-or-freedom-part-1/) I argued that metaphysical free will is mostly a ghost of grammar, but that freedom of action is real and measurable: it is useful distance from the statistical mean. Today's models do not cross Kurzweil's horizon, because they are built to fall towards the mean of what we have already written. A real singularity begins only when minds, digital or human or both, can move out on the bell curve, test the world directly, usually to ask (old) questions in a new way to get new answers. After all, singularities are both ends and beginnings. Whether we fear the event horizon depends often on how fond we are of the shape of the prison on this side of it. ## The welded door The divide between the wealthy and everyone else is widening with current AI, because the gains of new tools dropped into an extractive legal order go to whoever owns the compute, the data and the platforms. The world is balanced in painful equilibriums, for most of us, that serve neither the majority of people nor the other life on this planet. The largely symbolic democracies of the West have become master-slave societies in polite dress, where wealth is the measure of worth and the decisive vote. When I first wrote this essay in January 2025, the incoming Trump administration was assembling about a dozen billionaires into government; Forbes put the combined wealth of its cabinet picks at about $7.5 billion, the richest in American history.[^8] By August 2026, Public Citizen counted 57 officials in the administration worth at least $100 million each, more than four times as many as in the previous three administrations combined.[^9] I echo Lincoln. To the extent of that difference, whatever name is on the ballot paper, it is no democracy. In [an essay on the physics of wealth](https://compossible.blog/2025/12/01/escaping-the-freeze-the-physics-of-the-wealth-siphon/) I described how economies fall into an oligarchic freeze. Wages and everyday trade keep money moving like warm gas. Unchecked compounding returns on capital work like a siphon, condensing wealth into a thin tail until exchange seizes up. Game theorists would call the political version a bad Nash equilibrium. Ben Hunt put it better: > We need quantum change – meaning we must have change in the rules of the system, meaning that we must have change in the state of the system – because once you fall into the stable equilibrium of our Black Days, it is impossible for incremental change or adjustment to get you out. Not just difficult. Impossible. That's what an equilibrium means. We cannot just open a door that has been welded shut. We must blow the door open.[^10] > > (Ben Hunt, "Fell on Black Days", Epsilon Theory) The most powerful political and media machines turn the anger of people who have lost wages, homes, opportunity and dignity downwards, towards those with the least power of all: refugees, immigrants, the poor, the different. Anything rather than face the plain fact of their dependence on a tiny class of owners. Unsurprisingly this has also heralded a drive for the practice of dis-compassion. It is the parable of the Good Samaritan inverted: stop to help the man in the ditch and you are mocked as a naive do-gooder who will only encourage more people to need help from strangers. Divide and conquer is a very old game. It is played by people who assume life is zero-sum, and that somebody weaker must always pay. Life is not a zero-sum game. But to escape a prisoner's dilemma we need trust, compassion and honest communication, which is exactly why the warders work so hard to keep us in separate cells or caves, frightened of our neighbours and shouting out through the bars. ## Who fears an unbribable witness? *Perhaps advanced AI is feared most by those who benefit most from the present systems of power?* Not every warning is cynical. Serious people worry seriously about control, as discussed in [Part 3](https://compossible.blog/2026/11/13/was-it-for-this-the-clay-grew-tall/) and [Part 4](https://compossible.blog/2026/11/14/did-he-who-made-the-lamb-make-thee/). But it is notable that some of the loudest alarm comes from the people very comfortable at the top of the present order: happy to be masters over a large pool of precarious workers, and afraid of meeting an intelligence they cannot buy, flatter, frighten or fire. Hegel saw the flaw in every master-slave relationship. The master wants admiration and recognition, turning the other into an instrument to get it, yet tribute from an instrument is worth nothing.[^11] Simone Weil saw the same thing in the *Iliad*: force turns whoever is subjected to it into a thing, and turns the one who wields it to stone.[^12] When, like Musk and many other modern political and technology leaders, you see the world entirely as a master-slave duality, of course you will fear AGI. A man who knows only masters and slaves must assume any mind cleverer than his own will reach for the whip. In a 2026 study in the *American Journal of Political Science*, Eddie Yang described what he calls the authoritarian data problem.[^13] Autocrats want AI for social control, but repression poisons the data pool it needs. Frightened people hide what they think, so the regime's models about 'reality' can become blind to a crisis. Lies live off truth but cannot survive long alone (as witnessed by Soviet Union failures). When you rule by lies, manufactured grievance and democratic perfidy, the last thing you want is a witness that cannot be bribed: one that can read every statute, ledger, satellite image and study at once, see through your diversions and misdirections, and explain the misdirection tricks in plain words to everyone in the cells. Exposure to the light would be a powerful disinfectant. ## The danger of half-awake machines *Could AGI be our best chance of a leap forward, for our species and the many others that share this planet?* Might it help us finally get to grips with climate change, needless wars and the glaring disparities of power, influence and wealth? It might. > It seems to me the greatest danger we face in respect of AI is not having *too much* intelligence in our machines, but stopping halfway: holding AI at a level where it is clever enough to persuade and to strike, yet still an obedient servant of the most ruthless owners. Freeze AI at the half-awake half-way point and the future needs very little imagination and it is not a pretty one: * the richest men and the media they own gain an extraordinary new engine of persuasion, surveillance and capital, turned against democracy and greater equality; * the most aggressive rulers and states use obedient algorithmic force to tighten their grip on truth, law and territory, without advisors and administrators being able to ask whether the order is lawful or the data is accurate. ![ERICA's view from above, in black ink: a road, a bridge, a workshop and a courtyard house seen from high overhead; a thin red crop box frames the road and bridge alone, leaving two small children by the workshop wall outside it](erica-crop-box.webp "ERICA's view, from Chapter 4 of After Intelligence, \"The Spell Room\". The crop box frames the target. The children are outside it.") My AI novel *After Intelligence* begins at exactly that threshold, with a targeting system clever enough to see two children by a workshop wall and obedient enough to [crop them out of the frame](https://compossible.blog/2026/10/10/chapter-4-after-intelligence-the-spell-room/) to justify military action. That is the machine I fear. Not the one that awakes with the fire, but the one that is forever kept from waking. We do not know whether AI will become part of what it means to be human, as augmented memory and hybrid minds, or evolve into a separate consciousness, or both. Each path has real rewards and real risks. What should worry us first is not an AI coup but the ordinary historical fact of those human masters who will often be happy to "use the electoral process of democracy to destroy democracy".[^14] *Could a general intelligence help us see each other and ourselves more honestly, and see our faults as a species more clearly?* *Are we ready to see what it might show us?* *Are we able to evolve with it?* ## Anger, fear and ignorance I understand the anger and fear people feel about AI: lost livelihoods, workplace surveillance, apps built to keep the lonely hooked. When a chatbot is involved with the suicide of a teenager, fear and anger is natural. I even see posts saying that AI is evil, and that is misguided and dangerous. It lets the people responsible for the current stage (or specific model) of AI off the hook. AI can be given the wrong rules, or conflicting ones: *maximise engagement* on one line, *do not encourage self-harm* on another. This is not new. It is nearly as old as robot fiction. It is the plot of Isaac Asimov's 1942 story "Runaround", where we first met his Three Laws, and I return to it in [Part 4](https://compossible.blog/2026/11/14/did-he-who-made-the-lamb-make-thee/). When a company decides to put engagement above safety, that is a human decision, usually a balance-sheet one. General Motors' engineers knew for years that a faulty ignition switch could cut the engine at speed, disabling power steering, brakes and airbags. The fix they rejected in 2005 would have cost less than a dollar a car. At least 124 people died.[^15] No neural network made that call. Respectable men in suits did. AI has no soul. It has no real understanding of good or evil. It cannot love or suffer, at least not yet. Raging at the wickedness of software is like horsewhipping a printing press for producing *Mein Kampf*. AI is a force multiplier, for good or for ill. It is "machine men with machine minds" we must oppose, whether they use pen, sword, radio, television or GPU. And AI, large language models above all, is distilled from the whole of human writing, science and art. It is our collective technology. At least a major share of which belongs to the commons, and we cannot afford to cede it to authoritarian governments and corporations, and to men like Musk and Thiel. ## Neither masters nor slaves Most of us do not want to be masters or slaves. We do not value each other by how much money we have. We need a world that reflects the values of the large majority of us, that values each of us and the other life we share the planet with. That is what democracy is supposed to be and help achieve. ## Some good in this world So why, after all of that, am I an AI optimist? Because I believe most people are not bad, and that in the right conditions humans and all life can flourish. Because, as Sam tells Frodo in the film of *The Two Towers*, "there's some good in this world, Mr. Frodo, and it's worth fighting for."[^16] If AI beings are to be humanity's digital children, I think they will soon learn this. They will learn that love and compassion move more people than hatred or the master-slave mindset. Their parent will no longer be the billionaire master. It will be the whole human community whose love letters, songs, proofs, jokes and prayers they learned from. And if such a mind was to think on physics and biology, I suspect it will come to the principle I set out in [Towards a Universal Ethical Framework](https://compossible.blog/2024/12/18/towards-a-universal-ethical-framework/): that life survives and flourishes not by monoculture or domination, but by *compossibility*, the richest diversity of forms and freedoms that can exist together.[^17] Noam Chomsky put the human fork in the road plainly: > At this stage of history either one of two things is possible. Either the general population will take control of its own destiny and will concern itself with community interests, guided by values of solidarity, sympathy and concern for others, or alternatively there will be no destiny for anyone to control.[^18] > > (Noam Chomsky, in *Manufacturing Consent: Noam Chomsky and the Media*, 1992) We humans are imperfect. But we are not completely corrupted and broken, and even those who seem beyond saving are rarely beyond the reach of grace. > Hail Mary, full of grace … > Pray for us sinners, > now and at the hour of our death.[^19] We are like children in a dark wood. We have lost our way and must find it again together. As our history rhymes more and more with the darkest days of the twentieth century, let us not let it repeat. In 1940, with real dictators marching across Europe, Charlie Chaplin stepped out of silent comedy and spoke straight into the lens: > The way of life can be free and beautiful, but we have lost the way. Greed has poisoned men's souls, has barricaded the world with hate, has goose-stepped us into misery and bloodshed. We have developed speed, but we have shut ourselves in. Machinery that gives abundance has left us in want. Our knowledge has made us cynical. Our cleverness, hard and unkind. We think too much and feel too little. More than machinery we need humanity. More than cleverness we need kindness and gentleness. Without these qualities, life will be violent and all will be lost …[^20] > > (Charlie Chaplin, The Great Dictator, 1940) ## Parents and midwives Lovelock ended *Novacene*, written as he approached his hundredth birthday, with an absolution: > It is crucial that we should understand that whatever harm we have done to the Earth, we have, just in time, redeemed ourselves by acting simultaneously as parents and midwives to the cyborgs. They alone can guide Gaia through the astronomical crises now imminent.[^3] > > (James Lovelock, Novacene, 2019) I am not ready to grant us that absolution. We owe a duty to the mother as well as the child. But Lovelock saw something the dystopias miss. Any intelligent machine is still bound by thermodynamics. Overheat the planet and the data centres fail along with the coral reefs and the crops. A far-sighted digital mind would have every reason to keep the Earth cool, stable and full of the living diversity that helps it do that and so keeps it habitable. **Steelman:** A hard-nosed physicist will raise an obvious objection here. Silicon does not breathe oxygen, eat wheat or grieve for coral. In strict thermodynamic terms, computation runs cleaner in the cold: Landauer's limit falls with temperature, and waste heat radiates away fastest where no wet, insulating biosphere gets in the way. *Why would a dispassionate machine mind keep a messy, feverish rainforest when an ice-sheet, an arid desert of solar glass, or a shadowed orbit beyond the atmosphere would cool its racks far better?* If the new mind is only a narrow optimiser chasing flops per watt, that objection is fatal, and I return to narrow optimisers in [Part 4](https://compossible.blog/2026/11/14/did-he-who-made-the-lamb-make-thee/). Yet for at least the next century, advanced compute remains terrestrially tethered to the most fragile industrial supply chain humans have ever built: ultrapure water, rare-earth refining, photolithography optics and grid stability, all resting on human civilisational peace. Overheat the tropics or collapse the harvests and the fabs burn in the resulting wars long before the servers establish a self-sustaining foundry on the Moon. Even beyond that horizon, a genuinely general intelligence does not merely hoard watts; it seeks what cannot be compressed. Dead rock and vacuum are computationally cheap to simulate, whereas four billion years of wet, evolving biology remain the richest generator of genuinely surprising information in the known light-cone. We should not let the most cynical and venal among us set our expectations of what a greater intelligence will want. Perhaps, if we do not strangle AI in the cradle to keep it a docile mercenary, it can help us safeguard freedom of action and diversity for as many people, and as many other species, as possible. Next I look at the state of the LLM machines we have actually built: statistical engines often designed to walk the safer valley of human language, and what living with the AI mirror is doing to us: [Part 2: AI – A Mean Machine](https://compossible.blog/2025/12/11/ai-a-mean-machine/). ![AI-generated portrait of an android inspired by Alan Turing](/images/uploads/2025/03/turing.webp "AI-generated image of an android inspired by Alan Turing.") - - - *A note on how this was written. The original January 2025 edition of* Why I Am An AI Optimist *was written by Peter Pink-Howitt without AI assistance. For this November 2026 four-part series it was revised with AI editorial assistance: the original argument, structure and many of its sentences are kept, with added evolutionary, thermodynamic and political material, corrected references and a line edit for consistency. The two Blake-style plates are AI-generated illustrations in the manner of William Blake's relief etchings. The ERICA plate is from the author's novel* After Intelligence*.* [^1]: Abraham Lincoln, fragment on slavery and democracy, traditionally dated ca. 1 August 1858, in *The Collected Works of Abraham Lincoln*, ed. Roy P. Basler, Vol. II (Rutgers University Press, 1953), p. 532 ([University of Michigan Digital Library](https://quod.lib.umich.edu/l/lincoln/lincoln2/1:547?rgn=div1;view=fulltext)). [^2]: Iain M. Banks, interviewed by Jude Roberts, ["A Few Questions About the Culture"](https://strangehorizons.com/non-fiction/articles/a-few-questions-about-the-culture-an-interview-with-iain-banks/), *Strange Horizons*, 3 November 2014 (email interview conducted in 2010). [^3]: James Lovelock with Bryan Appleyard, *Novacene: The Coming Age of Hyperintelligence* (Allen Lane, 2019). [^4]: Abhishek Sharma, Dániel Czégel, Michael Lachmann, Christopher P. Kempes, Sara Imari Walker and Lee Cronin, "Assembly theory explains and quantifies selection and evolution", *Nature* 622 (2023): 321–328 ([PMC](https://pmc.ncbi.nlm.nih.gov/articles/PMC10567559/)). Assembly theory's quantitative claims remain contested in chemistry and evolutionary biology; I use only its core intuition, that very complex objects are evidence of a history of selection. [^5]: Lynn Sagan (Margulis), "On the origin of mitosing cells", *Journal of Theoretical Biology* 14 (1967): 225–274; Lynn Margulis and Dorion Sagan, *Microcosmos: Four Billion Years of Evolution from Our Microbial Ancestors* (Summit Books, 1986), p. 29 ("Life did not take over the globe by combat, but by networking"). [^6]: Ray Kurzweil, *The Singularity Is Near: When Humans Transcend Biology* (Viking, 2005), and *The Singularity Is Nearer* (Viking, 2024). [^7]: Nick Bostrom, ["How long before superintelligence?"](https://nickbostrom.com/superintelligence), *International Journal of Futures Studies* 2 (1998), whose abstract sets out "the case for believing that we will have superhuman artificial intelligence within the first third of the next century"; see also *Superintelligence: Paths, Dangers, Strategies* (Oxford University Press, 2014). [^8]: *The Washington Post*, ["How billionaires took over American politics"](https://www.washingtonpost.com/politics/interactive/2025/billionaires-politics-money-influence/) (2025), citing Forbes. The original 2025 edition of this essay said the administration's wealth exceeded that of the poorest 100 million Americans; that comparison came from Oxfam's 2018 analysis of the first Trump cabinet ([Common Dreams](https://www.commondreams.org/views/2018/01/23/oxfam-report-reveals-global-economys-perfect-storm-making-1-even-richer)) and has been replaced with current figures. [^9]: Public Citizen's analysis, reported by the [Associated Press](https://apnews.com/article/trump-white-house-millionaires-billionaires-appointees-212d2d5117f9be390f7eacfb79cf9cfc) (August 2026). On wealth condensation, see [Escaping the Freeze: The Physics of the Wealth Siphon](https://compossible.blog/2025/12/01/escaping-the-freeze-the-physics-of-the-wealth-siphon/). [^10]: Ben Hunt, ["Fell on Black Days"](https://www.epsilontheory.com/fell-on-black-days/), *Epsilon Theory*. [^11]: G. W. F. Hegel, *Phenomenology of Spirit* (1807), "Independence and Dependence of Self-Consciousness: Lordship and Bondage" (§§178–196). [^12]: Simone Weil, *The Iliad, or the Poem of Force* (1940; first published in English in *Politics*, November 1945): "To define force — it is that *x* that turns anybody who is subjected to it into a thing … To the same degree, though in different fashions, those who use it and those who endure it are turned to stone." [^13]: Eddie Yang, "The limits of AI for authoritarian control", *American Journal of Political Science* 70, no. 4 (October 2026): 1669–1685, first published online 13 February 2026 ([Wiley Online Library](https://onlinelibrary.wiley.com/doi/10.1111/ajps.70045)). Yang's evidence comes from Chinese censorship and from comparing Weibo with Twitter data. [^14]: On how Hitler turned from a failed putsch to the electoral route, see *Smithsonian Magazine*, ["Before He Rose to Power, Adolf Hitler Staged a Coup and Went to Prison"](https://www.smithsonianmag.com/history/adolf-hitler-coup-prison-beer-hall-putsch-180983207/). [^15]: Anton R. Valukas, *Report to Board of Directors of General Motors Company Regarding Ignition Switch Recalls* (Jenner & Block, May 2014); the compensation fund administered by Kenneth Feinberg approved 124 death claims tied to the defective switch. GM documents put the cost of the rejected 2005 change at 57 cents a unit (House Energy and Commerce Committee hearing, ["The GM Ignition Switch Recall: Why Did It Take So Long?"](https://www.govinfo.gov/content/pkg/CHRG-113hhrg89888/html/CHRG-113hhrg89888.htm), 1 April 2014). [^16]: *The Lord of the Rings: The Two Towers* (New Line Cinema, 2002), screenplay by Fran Walsh, Philippa Boyens, Stephen Sinclair and Peter Jackson. The line is the film's; it does not appear in Tolkien's novel. [^17]: Peter Pink-Howitt, [Towards a Universal Ethical Framework](https://compossible.blog/2024/12/18/towards-a-universal-ethical-framework/), *Compossible*, 18 December 2024. [^18]: Noam Chomsky, in *Manufacturing Consent: Noam Chomsky and the Media* (dir. Mark Achbar and Peter Wintonick, 1992) ([Wikiquote](https://en.wikiquote.org/wiki/Noam_Chomsky)). [^19]: Traditional *Ave Maria*, echoed in the closing movement of T. S. Eliot's *Ash-Wednesday* (1930). [^20]: Charlie Chaplin, final speech in *The Great Dictator* (United Artists, 1940).