Top Science Fiction and Fantasy Magazines 2026

Since 2014, I've compiled an annual ranking of science fiction and fantasy magazines, based on prominent award nominations and "best of" placements over the previous ten years. If you're curious what magazines tend to be viewed by insiders as elite, check the top of the list. If you're curious to discover reputable magazines that aren't as widely known (or aren't as widely known specifically for their science fiction and fantasy), check the bottom of the list.

Below is my list for 2026. (For previous lists, see here.)

Method and Caveats:

(1.) Only magazines are included (online or in print), not anthologies, standalones, or series.

(2.) I give each magazine one point for each story nominated for a Hugo, Nebula, Sturgeon, World Fantasy, Ignyte, or Eugie award in the past ten years; one point for each story appearance in the past ten years in the "best of" anthologies by Dozois, Horton, Strahan, Clarke, Adams, and Tidhar; and half a point for each story appearing in the short story or novelette category of the annual Locus Recommended list. New This Year: Since "best of" anthologies are becoming sparser, I've added the Ignyte and Eugie awards, including retrospectively to 2020 for the Ignyte and through the whole period for the Eugie.

(3.) I am not attempting to include the horror / dark fantasy genre, except as it appears incidentally on the list.

(4.) Prose only, not poetry.

(5.) I'm not attempting to correct for frequency of publication or length of table of contents. I've noticed that dividing by the number of stories generates results that fit less well with expert opinion. A magazine's reputation appears to be driven mostly by the visibility of the most visible stories, not by the visibility of the average story. (Similar remarks apply to the prestige of philosophy journals.)

(6.) I'm also not correcting for a magazine's only having published during part of the ten-year period. Reputations of defunct magazines slowly fade, and sometimes they are restarted. Reputations of new magazines take time to build.

(7.) I take the list down to 1.5 points.

(8.) I welcome corrections.

(9.) I confess some ambivalence about rankings of this sort. They reinforce the prestige hierarchy, and they compress complex differences into a single scale. However, the prestige of a magazine is a socially real phenomenon worth tracking, especially for the sake of outsiders and newcomers who might not otherwise know what magazines are well regarded by insiders when considering, for example, where to submit.

[Corrected 13:21 PDT: Fusion Fragment had been omitted due to a transcription error. Thanks to C.E. Singer for noticing the mistake.]

Results:

1. Reactor / Tor.com (193 points)

2. Clarkesworld (192.5) 

3. Uncanny (188)

4. Lightspeed (128) 

5. Asimov's (101) 

6. Fantasy & Science Fiction (95.5) 

7. Beneath Ceaseless Skies (63) 

8. Strange Horizons (incl Samovar) (55.5)

9. Analog (39.5) 

10. Apex (39) 

11. Nightmare (38.5) 

12. FIYAH (30) 

13. Slate / Future Tense (23.5) (ceased 2024) 

14. Fireside (20.5) (ceased 2022)

15. The Dark (18) 

16. Fantasy Magazine (16.5) (off and on during the period) 

17. The Sunday Morning Transport (14.5) (started 2022)

18. Interzone (13.5) 

19. Diabolical Plots (12.5)

20. The Deadlands (12) (started 2021)

21. khōréō (7.5) (started 2021)

22. Future Science Fiction Digest (7) (ran 2018-2023) 

22t. PodCastle (7)

24t. Conjunctions (6.5) 

24t. GigaNotoSaurus (6.5)

24t. Lady Churchill's Rosebud Wristlet (6.5)

27. The New Yorker (6) 

28t. Omni (5.5) (classic magazine relaunched 2017-2020) 

28t. Psychopomp (5.5) (started 2023; not to be confused with Psychopomp Magazine) 

30. Kaleidotrope (5) 

31t. Anathema (4) (ran 2017-2022)

31t. Baffling (4) (started 2020)

31t. Boston Review (4) 

31t. Omenana (4)

31t. Terraform (Vice) (4) (ceased 2023)

31t. Wired (4)

37t. B&N Sci-Fi and Fantasy Blog (3.5) (ceased 2019)

37t. Paris Review (3.5) 

37t. Shimmer (3.5) (ceased 2018)

37t. Sirenia Digest (3.5) 

41t. Science Fiction World (3)

41t. Galaxy's Edge (3) (ceased 2023)

43t. Augur (2.5) (started 2018)

43t. Beloit Fiction Journal (2.5) 

43t. Bourbon Penn (2.5)

43t. Buzzfeed (2.5) 

43t. Escape Pod (2.5)

43t. Flash Fiction Online (2.5)

43t. McSweeney's (2.5) 

43t. Tin House (2.5) (ceased short fiction 2019) 

*51t. Abyss & Apex (2)

51t. Black Static (2) (ceased 2023)

51t. Fusion Fragment (2) (started 2020)

51t. Mothership Zeta (2) (ceased 2017) 

*51t. PseudoPod (2)

51t. Reckoning (2)

51t. Shortwave (2) (started 2022)

*51t. Weird Horror (2)

*59t. Constelación (1.5) (ran 2021-2022)

59t. MIT Technology Review (1.5) 

59t. New York Times (1.5) 

59t. Translunar Travelers Lounge (1.5) (started 2019)

[* indicates new to the list this year]

--------------------------------------------------

Comments:

(1.) Beloit Fiction Journal, Boston Review, Conjunctions, McSweeney's, The New Yorker, Paris Review, Reckoning, and Tin House are literary magazines that sometimes publish science fiction or fantasy. Buzzfeed, Slate and Vice are popular magazines, and MIT Technology Review, Omni, and Wired are popular science magazines that publish a bit of science fiction on the side. The New York Times ran a series of "Op-Eds from the Future" from 2019-2020. The remaining magazines focus on the science fiction and fantasy (SF) genre or related categories such as horror or "weird". All publish in English, except Science Fiction World, which is the leading science fiction magazine in China, and Constelación, which was bilingual English and Spanish.

(2.) For a sense of recent momentum, it's also interesting to consider a three-year window. Here are those results, down to six points:

1. Clarkesworld (67)  

2. Uncanny (54.5) 

3. Reactor / Tor (42.5) 

4. Lightspeed (36.5)

5t. Asimov's (22.5) 

5t. Strange Horizons (22.5)  

7. Beneath Ceaseless Skies (13) 

8. The Sunday Morning Transport (9.5) 

9. Nightmare (9) 

10. FIYAH (8.5)

11t. Apex (8) 

11t. F&SF (8) 

11t. The Deadlands (8) 

14. Analog (7.5) 

15. Diabolical Plots (6.5) 

16. Fantasy Magazine (6) 

(3.) Other lists: The SFWA used to have a qualifying markets list of "pro" science fiction and fantasy venues based on pay rates and track records of strong circulation. However, that list has been discontinued. Submission Grinder is a terrific resource for authors, with detailed information on magazine pay rates, submission windows, and turnaround times.

(4.) Starting about 10-15 years ago, the classic "big three" print magazines -- Asimov's, F&SF, and Analog -- were slowly displaced in influence by the leading free online magazines, Reactor / Tor, Clarkesworld, Uncanny, and Lightspeed (all founded 2006-2014). Last year, Asimov's, F&SF, and Analog were all purchased by Must Read Books, and the transition has been bumpy, with author contract problems, delayed or skipped issues, and broken websites. It remains to be seen if the classic big three can stay viable in print format.

(5.) Standalone anthologies (traditionally in print format) have also been declining, probably for similar reasons as print magazines -- that is, the increasing consumption of free SF online. For example, in the most recent (2025) Locus recommended reading list only 8 of 99 (8%) recommendations in the novelette and short story categories are from anthologies, compared to 23/111 (21%) ten years ago (2016).

(6.) Academic philosophy readers might also be interested in the following magazines that specialize specifically in philosophical fiction and/or fiction by academic writers: AcademFic, After Dinner Conversation, and Sci Phi Journal.

AI and Consciousness

AI and Consciousness, hot off the press at Cambridge University Press and free online until Aug 26! Summary: Our most advanced AI systems might soon – within the next five to thirty years – be as richly and meaningfully conscious as ordinary humans, or even more so, capable of genuine feeling, real self-knowledge, and a wide range of sensory, emotional, and cognitive experiences. In some arguably important respects, AI architectures are beginning to resemble the architectures many consciousness scientists associate with conscious systems. Their outward behavior, especially their linguistic behavior, grows ever more humanlike. Alternatively, claims of imminent AI consciousness might be profoundly mistaken. Their seeming humanlikeness might be a shadow play of empty mimicry. Genuine conscious experience might require something no AI system could possess for the foreseeable future – intricate biological processes, for example, that silicon chips could never replicate. The thesis of this Element is that we don't know. Science lags, while engineering sprints ahead.

Here: https://www.cambridge.org/core/elements/ai-and-consciousness/E77C92088DA3C9F89E7FE7C75CBB1896

Obfuscatory Philosophy as Intellectual Authoritarianism and Cowardice (repost)

Aloha from Kauai! Here's another snail for your enjoyment.

Since I'm still on vacation, here’s a repost with edits from 2011 and from my 2019 book A Theory of Jerks and Other Philosophical Misadventures.

[giant African land snail, seen in southern Kauai]

I've been told that Kant and Hegel were poor writers whose impenetrable prose style is incidental to their philosophy. I've also been told that their views are so profound as to defy expression in terms comprehensible even to smart, patient, well-educated people who are not specialists in the philosophy of the period. I've heard similar things about Laozi, Heidegger, Plotinus, and Derrida. (I won't name any living philosophers.) I don’t buy it.

Philosophy is not wordless profound insight. Philosophy is prose. Philosophy happens not in numinous moments of personal genius, but in the creation of mundane sentences. It happens on the page, in the pen, through the keyboard, in dialogue with students and peers, and to some extent but only secondarily in private inner speech. If what exists on the page is not clear, the philosophy is not clear. Philosophers, like all specialists, profit from a certain amount of jargon, but philosophy need not become a maze of jargon. If private jargon doesn’t regularly touch down in comprehensible public meanings, one has produced not philosophy, but only hazy words of indeterminate content. There are always gaps, confusions, indeterminacies, hidden assumptions, and failures of clarity, even in the plainest philosophical writers, like Mozi, Descartes, Hume, and David Lewis. Thus, these philosophers present ample interpretative challenges. But the gaps, confusions, indeterminacies, and hidden assumptions lie only one step behind the visible prose, available to anyone who looks conscientiously for them, not sliding away into a nebulous murk.

For a philosopher who can convince readers to take him seriously (given the exemplars, let’s say “him”), obfuscation yields three illegitimate benefits. First, he intimidates the reader, and through intimidation, dons a mantle of undeserved intellectual authority. Second, he disempowers potential critics by having a view of such indeterminate form that any criticism can be attributed to misinterpretation. Third, he exerts a fascination on the kind of reader who enjoys the puzzle-solving aspect of interpretation, thus drawing from that reader more attention than the quality of his ideas may merit (though this third benefit may be offset by alienating readers with a low tolerance for lack of clarity).

Philosophers of this type exhibit a kind of intellectual authoritarianism, with themselves as the assumed authority whose words we must spend time puzzling out. And simultaneously they lack intellectual courage: the courage to make plain claims that could be proven wrong, supported by plain arguments that could be proven fallacious. These three features synergize: If a critic thinks she has finally located a sound criticism, she can be accused of failing to solve the fascinating interpretative puzzle of the philosopher’s superior genius.

Few philosophers, I suspect, deliberately aim to be obfuscatory. But I am inclined to believe that some are attuned to its advantages as an effect of their prose style and for that reason don’t bother to write comprehensibly. Maybe they find their prose style shaped by audience responses: When they write clearly, they are dismissed or refuted; when they produce a fog of words that hint of profound meaning underneath, they earn praise. Maybe they are themselves to some extent victims — victims of a subculture, or a circle of friends, or an intended audience, that regards incomprehensibility as a sign of brilliance and so demands it in their heroes.

The Summer Illusion (repost)

Aloha from Honolulu! I have vacation brain, so I'll do a summer-themed repost.

[Photo from the Bishop Museum malacology exhibit]

*************************************

Every spring I suffer the Summer Illusion. The following three incompatible propositions all seem to me, in the spring, to be true:

(1.) When summer arrives, I'll finally get a bunch of that research done which has been crowded out by my teaching and administrative commitments during the school year.

(2.) When summer arrives, I'll finally get a chance to do all of that non-academic stuff that I've been putting off during the school year -- big home maintenance projects, vacation travel to the four new places I want to visit, my plan to catch up on the whole history of golden-age science fiction.

(3.) When summer arrives, I'll finally have a chance to spend a lot more time just relaxing.

The Summer Illusion is surprisingly robust. Every spring, I suffer the Summer Illusion, building up big plans and hopes. Then, every summer, as those hopes fall apart, I scold my springtime self for having fallen, yet again, into the Summer Illusion. The pattern is so common and predictable I've given it a memorable name, The Summer Illusion, to help convince myself that it really is an illusion -- and hopefully not fall into it again. And yet I fall into it again.

You might think that the Summer Illusion depends on entertaining only one of the three propositions at a time. You might think that the way it works is that sometimes I entertain proposition 1 (I'll get my research done!), and at other, different times I entertain proposition 2 (I'll get all my other projects done!), and at still other times I entertain proposition 3 (I'll finally have lots of time to relax!). Largely this is so. And yet the Summer Illusion also survives simultaneous consideration of the three propositions. Even looking at the propositions side by side like this, I am tempted to believe them. Some part of me thinks of course all three can't be true, as I've seen time and time again -- and yet in my heart I continue to believe. Summer days expand so magnificently to fit my fantasies!

It's almost an inversion of busyness. If a period of time has the outward appearance of being a "relaxed", low-commitment period of time, it serves as a fantasy-and-procrastination magnet. I pile my future plans and hopes into that period of time, not noticing the impossibly mounting sum of expectations.

The Cognitive Advantages of Being of Two Minds, or Twelve, or Twelve Hundred

I was of three minds,
Like a tree
In which there are three blackbirds.

[from Wallace Stevens, Thirteen Ways of Looking at a Blackbird]

The literature on superintelligence and AI alignment tends to assume that an intelligent AI will have one set of values, one reward function, and consistent, determinate preferences. But speaking for myself, I'm not as consistent as that. You aren't either, I suspect. Now, maybe this is a defect in us that would be absent from a well-engineered intelligent system -- but I don't think so. There are advantages to having, so to speak, a splintered mind.

Two Minds for Two Theories

Suppose you're 85% confident Theory A is true and 15% confident Theory B is true. And suppose it's not obvious what follows from the two theories; discovering the implications will require substantial cognitive work. With that work, you could learn that Theory A implies C, D, and E, while Theory B implies F, G, and H. Here are three cognitive strategies:

All-in on the most probable. Assume Theory A. Work for a while and discover that C, D, and E follow. Declare yourself done.

Both in a single model. Build a single cognitive model with the representation .85*A & .15*B. Work with this complex representation and eventually discover .85*C & .85*D & .85*E & .15*F & .15*G & .15*H.

Two models. Assume Theory A. Work a while and discover that C, D, and E follow. Then assume Theory B. Work a while and discover that F, G, and H follow. Now weight the theories and their implications .85 and .15.

All-in is easiest and cheapest. But if you discover that Theory A is wrong and then need to act quickly on the consequences of Theory B, you'll be sorry you hadn't worked out those consequences in advance.

Single model is probably hardest and slowest. It's difficult to keep track of complex conjunctive representations.

Two models has all the advantages of single model, while being easier, less prone to error, and probably faster. I'll bet it's what you did when I described single model. You thought through the conjuncts separately, considering first what follows from A, then considering what follows from B, then stitching the results together, rather than thinking truly conjunctively through the whole process.

Here's something that might be even better, if you could do it: Explore the two models simultaneously in parallel. One part of you works out the implications of Theory A, while another part of you works out the implications of Theory B. If we have design choices in creating an intelligence, maybe we should set it up that way. Arguably, much of the human mind does work that way. Even if attentive conscious processing is narrowly focused and serial, parallel nonconscious processes also operate simultaneously behind the scenes (compare Dennett's characterization of cognitive processes competing for "fame in the brain").

If I'm building a mind that I want to be robust against error while working efficiently, I might encourage it to divide itself into subminds with different opinions. The submind with the likeliest opinion gets the most control over the behavior and final judgments, but especially in risky situations, the subminds might compromise. For example, if the Theory A submind recommends Action X as slightly better than Action Y, while the Theory B submind predicts that Action X would be catastrophic, the system as a whole might choose Action Y.

Obviously, this generalizes to numbers above two.

Subminds for Plans, Values, and Preferences

The same logic applies to plans. Mind 1 develops Plan A, while Mind 2 develops Plan B. If Plan A fails, Plan B is ready to go -- especially important if quick action is needed and Plan B takes time to construct.

Perhaps less obviously, values and preferences might also be divided. Mind 1 values hard work and exercise. Mind 2 values lolling about at home reading fiction. You're a morning person: You wake with energy, and Mind 1 is in charge. Off you go for a jog, then to the office, blazing through your to-dos. By late afternoon, your energy flags and Mind 2 starts taking over. You can't just let Mind 1 be always in charge. You'd exhaust yourself, and you'd become a one-dimensional workaholic. Similarly, you can't just let Mind 2 be always in charge. You'd lose focus and vegetate.

Hypothetically, a single, complex value representation might suffice -- the value equivalent of the "single model" above -- one stable, intricate set of conditionalized values, producing different behavior in the morning and evening.

But the system needn't be built that way. And, phenomenologically, that's not how it feels to me. It feels like my morning self simply values one set of things while my evening self simply values different things. Sometimes I wish they would get along better. In the morning, I resolve to eat healthily all day, and I succeed until 9:00 pm. Then my evening self concocts an excuse to eat three big cookies. The next morning, looking at the bathroom scale, my morning self excoriates my evening self.

Of course this is a simplification. The values are complex, both morning and evening. The point is, they're not constant. There's not, it seems, one steady underlying valuation. My cares, my preferences, my priorities fluctuate over time, depending on many factors -- not just time of day, but mood, situation, bodily state, what's salient at the moment, what memories or habits happen to arise. There isn't a stable preference ranking, a single momentary snapshot, that represents the real, undistorted me. My morning and evening and other selves are equally authentic and equally me.

As in the Theory A vs Theory B case, we can shift from diachronic to synchronic -- from serial to parallel, from this-then-that to this-and-that simultaneously. My work self, my family self, my friend self, with their different preferences and values, are all simultaneously ready and updating in the nonconscious background. My work self, or selves, dominate at the office, but I could swiftly transition if an old friend were unexpectedly to knock on the door. And once in a while these other selves toss "off topic" thoughts up into my central conscious stream.

Back to AI

Maybe this isn't how the human mind works. I'm inclined to think it is, but set that aside. We certainly could build AI to work that way. Multiple competing models of the world. Multiple competing value sets. Shifting with evidence, shifting with circumstance -- perhaps not even capable of being distilled into a single conjunctive representation or single coherent value ranking.

Multiply the minds -- generalizing from two to twelve or twelve hundred -- and the subminds can form coalitions. Mind 1 might have the plurality of the weight, but if Minds 2 to 5 all agree on something else, they might collectively outweigh it. With enough diverse subminds, the group dynamics of voting, coalition building, and delegation might arise.

Let's not assume that an AI superintelligence (if the concept even makes sense) would be a hyper-coherent, hyper-rational entity that never contradicts itself and always maximizes the same fixed set of values. It might be instead a swirl of many contradictory, competing processes that compromise among one another and take turns in control.

What would it feel like to be such an entity, if such an entity is capable of having experiences? Maybe it would feel something like being us. But I can also imagine such an entity being more multi-centered than we are (we, with our narrow, serial attention in a single stream of consciousness [or so we ordinarily assume]). Such an entity might have many eddies of consciousness, partly distinct, partly overlapping, not easily divisible into discrete, countable units -- neither one unified person, nor many discrete persons, but a swarm of partly discrete entities, like a network of half-joined plants or fungi that fade one into the next without sharp boundary.

[image source, NASA]

AI Slop and Evidence about Evidence: Why Philosophy Journals Should Reject AI-Written Prose

I don't want to read your AI-generated email. Lots of reasons why, but here's the core of it: I want the text to reflect your thoughts. I want to know that you, the human, actually had those ideas and had them authentically enough to express them in exactly the form I see on the page. For personal emails, I want this for personal reasons. For philosophically substantive emails, I want this for evidential reasons. Philosophy journals should also want human-generated rather than AI-generated text for the same evidential reasons.

There Probably Are Good Reasons You Phrased It the Way You Did

The evidential reason: Human experts think differently and better than LLMs. Their word choices, even subtle ones, reflect sensitivities that they might not themselves be aware of. Typically, an expert's prose will be more sensitive to the matters on which they are expert than the output of a language model. When I receive a philosophical email from you -- and more so when I read a journal article -- I want your expert word choices, not blurry LLM approximations.

You might object as follows: Of course I read the LLM outputs before sending, and I wouldn't send the email, much less submit the article, unless I endorsed every word! So, the objection continues, you did think the thoughts expressed. The text reflects your expert best judgment -- maybe even something better than your expert best judgment: your expert best judgment combined with the expertise of an LLM.

I reply: There's a huge cognitive difference between nodding along while reading something and actually productively generating a text. Two reasons: First, once the text is on the page, it's easy to passively let the approximate word suffice, rather than thinking about word choice in the same effortful, active way we do when generating prose de novo. Second, as I suggested above, I doubt that human beings, even experts, have a good sense of all the factors that shape word choice -- everything they're being sensitive to. You would have phrased it slightly differently, and even if you don't know that, or why, a different signal is sent and received.

Evidence about Evidence

I'm talking about evidence about evidence: meta-epistemology. Your email or your article presents evidence for a particular philosophical view (alternatively, evidence that you support a particular philosophical view). In a simple world, I could evaluate this evidence entirely on its face: How good is the proposed view? But in the actual, complex world, it helps to have evidence about the quality of the evidence. The fact that you, an expert human, generated the text is evidence that the view is worth thinking about -- more so than if the text were generated by an LLM. This holds even if the text is exactly the same, which of course it wouldn't be.

An increasingly large part of the function of journals is to provide evidence about evidence -- the value of their imprimatur. The fact that an article appears in Nous or Ethics is evidence that it has been through rigorous review and was judged worthy by several expert humans applying unusually demanding standards of quality and importance. Its appearance in those journals is thus evidence (imperfect of course!) that the reasoning is of high quality and the arguments worth taking seriously.

Similarly, if I know that an email or an article was written by a respected colleague, reflecting their positive creative exertion in trying to choose the right words, guided by their intuitive expertise in how to phrase things, I have better reason to take it seriously than if I know that it was generated by an LLM and reflects only their passive after-the-fact assent.

Philosophers sometimes suggest that we shouldn't care if an argument was human-generated or AI-generated -- that insisting on human-generated prose is fetishizing personal human interaction rather than facts and argument quality. In a way, that's true: A sound argument is a sound argument. Similarly, we shouldn't care if an article was written by David Chalmers and published in Philosophical Review or whether it was written by someone with no institutional affiliation and published on an obscure blog. If the argument is good, it's good -- of course, of course!

But at the same time, we have limited attention, limited time, limited ability to understand the nuances when matters drift even a little from our tightest foci of expertise, and in these cases it's helpful to have meta-evidence. What should I read? How far should I trust the author has the details right, versus how much should I pause critically and chase down independent sources? How much should I let their way of phrasing things, their habitual patterns of thinking, the presuppositions hidden in their word choices and sentence structures, slip gently into my brain, silently strengthening my own associations and predilections?

An Example

To better capture the phenomenon I have in mind, I've selected a random sentence from my most recent outbound email (as of the first draft of this post):

This lack of self-respect is manifested both in her willingness to consider being killed to somewhat improve Josie's life and in her incapacity to develop desires and interests of her own independent of Josie's, as manifested by her being content to rot in a junkyard once her usefulness to Josie has ended.

The topic is Klara from Klara and the Sun, the story of a robot servant who is designed to be utterly dedicated to serving the human girl to whom she has been assigned as an "artificial friend". (I discuss the case in detail here.)

I asked ChatGPT "high" to "write the sentence better", and here is the output:

This lack of self-respect is evident both in her willingness to contemplate her own destruction for the sake of a modest improvement in Josie’s life and in her failure to develop desires or interests independent of Josie’s—as shown by her contentment to decay in a junkyard once she is no longer useful to her.

At first glance, the revised sentence is slightly better -- smoother, more natural. My double use of "manifested" was awkward, and rereading it now I think I should have changed at least that stylistic detail. I could even see approving the LLM output as expressing my own thinking about Klara, if someone asked whether it accurately reflects my view.

However, in several subtle ways the edited version is philosophically worse.

I wrote that Klara's lack of self-respect is "manifested" in her willingness to such-and-such. ChatGPT replaced "manifested" with the slightly more common and natural "evident". Close enough, seemingly? But no. They differ subtly, in an important way. "Manifested" is ontological, "evident" is epistemic. "Manifested" emphasizes the relation between the underlying lack of self-respect and the particular thoughts Klara is having -- that those thoughts arise from a broad insufficiency of self-respect. "Evident" emphasizes that the particular thoughts are evidence of that attitude. Either word might do, but "manifested" is better: I'm not really talking about what evidence we have that Klara lacks self-respect but rather what types of thinking constitute lack of self-respect.

The second change: "consider being killed" is replaced by "contemplate her own destruction". The LLM phrasing is again a bit smoother, more common, more natural. And again, it's not exactly wrong; it's a phrasing I might easily approve. But on careful thought, my original phrasing is again truer to what I really want to say. "Consider" suggests that Klara is seriously weighing the choice. "Contemplate" is vaguer: She might only be thinking idly about it. "Being killed" is vivid and suggests that someone else would be the agent of her death. "Her own destruction" is not as vivid: "Destruction" flattens things a little, morally and emotionally. And it's not quite as clear that someone else, rather than Klara herself, would be the agent of her death.

And so I could continue with the other revisions.

I am focusing on line-by-line prose, not main ideas. Main ideas are a different issue and require a different analysis. One might wrongly think that only the main ideas are important. My point here is that subtle differences in line-by-line prose matter too, and that an LLM's word choices will differ from an expert's, and that there's reason to respect experts' intuitive word choices. If experts yield their writing to a language model, it is probably too easy for them to allow their expert word sensitivities to be blurred away.

Thus, people should prefer human-expert-generated prose in emails and journal articles, for at least this meta-epistemic reason. That a passage was written by an expert human is evidence that the prose accurately tracks important nuances. Being hand-written by a human expert is an (imperfect) high-level indicator of quality.

Those of us who receive emails, or read journal articles, or even who edit and referee journal articles cannot evaluate every sentence with the same close attention to nuance I demonstrated in my example above. We often need to read at two minutes a page, not two minutes a sentence. We all reasonably rely to some extent on the writer's expertise in phrasing.

This is one reason journals should require human-written rather than AI-written prose, even if the latter is endorsed post-hoc by a human expert.

Using LLMs Well

This is not a call for blanket rejection of the use of LLMs in philosophical (or other) writing. They can be useful for generating criticisms and for suggesting topics and readings to explore, as long as one takes the outputs with a cautious grain of salt.

They can also be useful for copyediting after the initial prose is written -- but only either

(a.) for minor grammatical/spelling corrections, or

(b.) in a thoughtful, effortful way, ideally with mechanisms to reduce passivity such as (b.i) having to type in any adjustments by hand rather than simply approving them in a revised document and (b.ii) receiving suggestions from more than one LLM to force you to contemplate several alternatives rather than simply accepting one, and

(c.) in a way that preserves or amplifies, rather than averages away, your distinctive way of expressing yourself.

Such an approach to LLM-aided revision is time-consuming rather than time-saving, a matter of "cognitive onloading" rather than "cognitive offloading" -- an occasion to actively consider your word choice line by line, via input from a few not-very-insightful copyeditors. Had I subjected my email to LLM copyediting, I'd have noticed the awkward double use of "manifested" and consequently considered whether that was the best way to express myself. The result of AI-assisted copyediting following principles (a)-(c) might be prose that even more accurately reflects your best individual thinking than your independently produced drafts.

This argument against using LLMs to generate expert prose is bounded and possibly temporary: If someday LLMs surpass human experts at subtle word choice, this particular argument will no longer apply -- though other excellent reasons might remain not to use LLMs for expert prose writing.

[Mud: image source]

New Book in Draft! Humanlike: A Defense of AI Rights

Here. Comments welcomed, hoped for, treasured.

Over the past couple years I've been working on a pair of books, AI and Consciousness and Humanlike: A Defense of AI Rights. I began circulating AI and Consciousness last fall, and it should appear with Cambridge Elements soon. Both experts' and non-experts' comments were extremely helpful in revising. Thanks so much!

I'd like to similarly begin collecting comments on Humanlike, starting today. Any reader who gives comments on the whole book will receive an appreciatively signed copy when it appears in print, as well as (of course) a call-out in the acknowledgements section.

To those wondering about the apparent madness of working on two books at once: They're both short (30K and 50K words, respectively) -- really one good-size book in total. Also, more than half of Humanlike is synthesized and updated material from several published and forthcoming articles going back to 2015.

I see the books as a companion pair. AI and Consciousness is a skeptical overview of the cases for and against AI consciousness. I argue that neither the boosters nor the scoffers have compelling arguments. Consequently, we will probably soon (within five to thirty years) have AI systems who might be as richly conscious as human beings or that might be as experientially blank as toasters. We won't have good scientific or philosophical grounds to settle the question.

Humanlike explores the ethical consequences. The central theses are:

(1.) AI with humanlike consciousness would deserve humanlike rights.

(2.) AI whose humanlikeness is seriously debatable should not be created, since it will force us into a dilemma between possibly overattributing and possibly underattributing rights, with potentially catastrophic consequences either way.

(3.) Our intuitive and theoretical understandings of how to treat persons ethically, grounded as they are in a narrow range of familiar human cases, are likely to fail catastrophically when confronted with future AI "persons" with radically different lifeways -- for example who can divide, merge, overlap, and back themselves up.

We are as unready for conscious AI systems as medieval physicists were for spaceflight. Still, the eventual result of technological development, perhaps in the thousand-plus-year future, might be a planet so richly full of diverse sources of awesomely valuable existence that it resembles a new Cambrian Explosion.

Full text here.

Bare Functionalism Versus Digital Computational Functionalism

In philosophy of mind, people sometimes say things like the following: If functionalism is true about consciousness, then digital computers could be conscious; all it requires is that they instantiate the right programs. And yes, most functionalists about consciousness do think that. But the claim about digital computer programs does not straightforwardly follow from functionalism alone. Today I want to clarify the issue by distinguishing three forms of functionalism: bare functionalism, computational functionalism, and digital computational functionalism.

Bare Functionalism

According to bare functionalism -- that is functionalism, without further commitments -- mental states are functional states. They are determined wholly by their causal relations to stimuli, behavior, and other mental states. Whatever plays the causal role of some mental state M is mental state M. Whatever plays the causal role of pain, for example, being apt to be caused by tissue damage and tissue stress and being apt to cause writhing, groaning, complaint, avoidance, calls to the doctor, and angry vows of vengeance (to simplify somewhat) is pain -- whether it is brain state 1117A in humans, pod state 24uw in Martians, or computational state 0110100110 in a robot. It's also generally assumed that mental state terms are in principle eliminable en masse by Ramsification. That is, they are placeholders for whatever Xs, Ys, and Zs are caused by the specified stimuli and cause the specified behavior and have the right causal relations to other Xs, Ys, and Zs.

Thus, if two systems behave exactly alike given the same patterns of environmental stimuli and do so by going through the same types of abstractly specifiable internal state transitions, they are necessarily mentally identical, whether they are made of silicon, carbon, or plaster of paris. Of course some materials -- such as plaster of paris -- probably can't support the causal complexity of humanlike behavior, no matter how cleverly arranged. So that puts a practical limit on the feasible materials out of which a humanlike mind could be constructed. But the only constraint on the substrate or material underlying mentality is that it supports the relevant causal relations.

Of course, no two different materials will respond in exactly the same way to every perturbation. Functionalist substrate flexibility thus requires constraints on what count as relevantly different stimuli and relevantly different behaviors. Functionalists tend not to commit to specifics, but hand-waving generalizations convey the spirit: If we built a robot -- like one of Asimov's robots or Data from Star Trek -- who responded in broadly humanlike ways to broadly humanlike stimuli, and did so by means of internal mechanisms sufficiently like ours at an abstract causal level, the robot would have mentality like ours, whatever its underlying material constitution.

How similar must the internal mechanisms be? A single giant search tree or lookup table is probably too different. But suppose the subsystems segment the visual world into discrete objects, and store long-term memories, and store lexical items flexibly combinable via generative grammar, and monitor its bodily condition, goals, and mental states, and so on, for all of our most important cognitive functions abstractly described (or less question-beggingly, they do the abstract causal equivalent of all of these things). That's probably close enough, even if the details differ.

Computational Functionalism

Computational functionalism is stronger -- that is, riskier and more committal. It holds that functional computational structure alone suffices for consciousness. Bare functionalism needn't say this. Bare functionalism can require that stimuli, behavior, and internal causation be specified noncomputationally. Two important differences follow.

First, anti-computationalists, such as Searle, observe that a computational model of a hurricane gets no one wet and a computational model of an oven cooks no turkey. Unlike the computational functionalist, the bare functionalist can join Searle in insisting on wet turkey: To fully match the relevant input-output relations, the system much have certain concrete effects.

It might help to remember that functionalism is essentially a complexification of dispositional or causal approaches to properties. Just as we can define a poison as something that (simplistically) tends to sicken or kill when ingested, a functionalist can define a painful experience as something that tends to produce certain concrete behaviors, such as (simplistically) screaming, protest, and avoidance. If there's no tendency to concretely sicken, there's no poison; if there's no tendency to concretely scream, protest, and avoid, there's no experience of pain.

The computational functionalist rejects this commitment to specific, concrete inputs and outputs. As long as the right computational relations are instantiated, between some set of inputs and some set of outputs, I1...Ii and O1...Oj, abstractly described, mentality is present. This is because exactly the same computations have been executed, whatever the concrete Is and Os are.

Asimov's robots can throw water balloons and cook turkeys, and a functionalist who is not a computational functionalist can insist that this matters.

Second, although the bare functionalist holds that mental states depend on causal relations, unless pancomputationalism is true, not every causal relation is computational. On Piccinini's account of computation, for example, computation only occurs when "medium independent vehicles" -- physical variables defined solely in terms of their degrees of freedom (e.g., 0 vs 1) rather than their specific physical composition -- are manipulated according to rules. The bare functionalist can insist that some noncomputational causal relations matter to mentality.

Digital Computational Functionalism

Digital computational functionalism is stronger still. It further commits to treating the relevant computation digitally. Standard theories of computation, such as Turing's, are digital, and the best-known computers are digital computers. But other forms of computation are sometimes described and implemented, including analog computation and nondigital forms of computation inspired by, and perhaps instantiated in, the human brain.

Why This Matters

Keeping these distinctions clearly in mind can help us appreciate that functionalism is a larger tent than sometimes assumed. Functionalists don't need to think that digital computers could achieve consciousness. Functionalists can insist on genuine embodiment. Functionalists needn't insist that only abstractly specified causal relations matter. They do have to insist that what matters internally, among the Ramsifiable mental state relations, are the abstractly specified causal relations. But they can simultaneously insist on particular types of concrete inputs and outputs.

One might think this is an inelegant or flawed position, not the most natural development of functionalism; but that requires argument.

[COBOL Rube Goldberg by Philip Manker: image source]

New in Draft: Strange Intelligence: Moral Puzzles of Unhumanlike AI

Available here.

Abstract: Future AI persons might both (1.) deserve moral consideration and rights fully equal with natural human persons, and (2.) have lifeways so radically different from ours as to break familiar patterns of moral thinking by violating our ordinary background assumptions. This article presents a series of thought experiments about strange AI persons, centering on a two-pronged worry featuring two types of "monster". "Utility monsters", who derive great personal benefit from harming others, create a well-known challenge for ethical systems that aim to maximize aggregate goods. The less-discussed case of "fission-fusion monsters", who can divide and merge at will, presents a complementary challenge to ethical systems focused on individual rights, since individual rights frameworks require the existence of stable, countable individual persons. AI cases dramatically expand the range of possible lifeways, creating untested problem cases for ethical systems that assume persons of the familiar humanlike sort.

To focus on the issues of interest, I assume the possibility of AI consciousness in this article. My skeptical overview of the issue of AI consciousness is here.

Excerpt (with light modification for independent readability):

3. Fission and Identity.

Backup is only the most modest duplicative possibility. If backup is possible, duplicative fission almost certainly will be possible too. Buy the new robot body before the old one dies and install the "backup" right away. Now Shriya-1 and Shriya-2 exist contemporaneously -- twin sisters, so to speak, who begin even more identical than "identical" human twins. We might imagine a billionaire Shriya creating thousands of duplicates of herself -- maybe millions or billions, if expensive robot bodies are unnecessary. Directed or random variation might be introduced, blurring the line between duplication and new creation.[1]

Suppose that AI children are ordinarily born as follows. Two adult AI persons, such as Shriya and Alaleh, jointly create an immature infant AI in a blank robot body. The infant's initial parameters blend Shriya's and Alaleh's initial parameters, with some random variation or directed tweaking.[2] Under Shriya's and Alaleh's care, the infant slowly matures. Ordinary AI birth would then be very different from duplication. We can also imagine intermediate cases. Maybe there's a library of successful toddler-equivalent and adolescent-equivalent AI models from which prospective parents can choose. They can then add variation, whether random, eugenic, or inspired by their own features. (Let's not enter here into the hazards and moral puzzles of eugenics, which could easily fill a small library.[3]) Duplicating one's current AI self thus constitutes one end of a continuum of AI creation from infancy to maturity.

If Shriya-1 creates a virtually identical contemporaneous copy, Shriya-2, she has now, it seems, entered a polyamorous relationship with Alaleh. Shriya-1 and Shriya-2 will soon diverge. Maybe Shriya-1 works as a scientist every weekday, while Shriya-2 stays home with their newborn.

If Shriya-1 deserves rights, Shriya-2 seemingly deserves similar rights, despite her technically younger age. We wouldn't want people creating oppressed duplicates of themselves. We wouldn't want Shriya-1, for example, who loves science and hates housework, to create a miserable homemaker duplicate who can't strike out into an independent life.[4]

Maybe, probably, half of Shriya-1's money should go to Shriya-2, even though Shriya-2 is a newborn duplicate. Maybe, probably, Shriya-2 deserves just as much right to rescue, healthcare, legal protection, free speech, free movement, privacy, and legal contracts. Should Shriya-2 be a citizen? If she is stateless and voteless, she's not fully equal with Shriya-1.

But if Shriya-2 is a citizen and can vote, there's potential for abuse if some AI persons can create many duplicates. Suppose a wealthy Robo-Elon creates a million AI duplicates just in time to register for the November elections. To prevent such abuses, we might impose a waiting period before voting, though eighteen years seems excessive if the AI systems are already cognitively mature adults. More moderate waiting periods -- say, seven years, a typical waiting time for immigrants to apply for citizenship -- could still generate political chaos after a few election cycles.

Nor do the political problems stop with voting. Suppose Robo-Elon creates a million duplicates the day before the census. Or suppose that Robo-Elon's descendants apply for healthcare subsidies, unemployment benefits, enrollment in community college, and tours of the state capitol. We must either risk chaos or treat them worse than they seem to deserve.

Could we limit fissioning?[5] Maybe every AI person can fission only once per year, reducing tactical fission. But even at that rate, the AI population could double every year -- up to a thousandfold increase in a decade. In humans, pregnancy is a burden, babies are a lot of expensive work, and babies can't have their own babies for at least another 15-20 years. One solution -- though it might seem needlessly restrictive to the AI persons -- might be to enforce humanlike costs and delays. This approach handles the moral puzzles by designing AI systems to have humanlike reproductive lives, so that they fit smoothly into our existing institutions and understandings: See the Policy of Humanlike Design in the concluding section of this article.

Death again presents conceptual challenges. Suppose Shriya-2 dies the next day. This seems much less tragic than the death of an ordinary unduplicated, un-backed-up human being. But as she lives on, diverging from Shriya-1, her death becomes more significant. Her memories, values, skills, habits, and personality are changed by living as a homemaker, raising an AI infant, until she becomes very different from Shriya-1, who works at the lab late into the night. Again we face the Death Dilemma: Either retain a sharp-edged metaphysics of death and lose much of death's moral significance or retain the moral significance and treat "death" as a matter of degree.

How deathlike is the death of a backed-up or duplicated AI? Maybe it depends on the age of the backup or the time since duplication, the fidelity of the backup or duplicate, and the time and changes accumulated as an independent entity. One possibility: These factors all reduce to a common factor of difference between the dying person and the backup replacement or duplicative alternative. The greater the difference, whether due to time or infidelity, the more deathlike the death.

Or maybe independent existence carries its own weight, in addition to difference? Suppose two duplicates split twenty years ago but retained virtually identical personalities and lived virtually identical lives, perhaps making similar decisions in parallel virtual realities. The pure accumulation of time, and of relationships to different persons and events, however similar, might make the death of one of them much like ordinary human death despite their similar features. After all (arguably) the spouse of Person A loves specifically Person A and not some other person, however similar. Their beloved, specifically, has died. Can we separate the importance of simply living a life over time from the importance of having different relationships to people and events, which cease upon death?

Might the ethics depend on the purpose for which the duplicates are created and their own attitudes toward "death"? Robin Hanson imagines people duplicating themselves to make decisions.[6] If you can't decide where to go to college, or what stocks to buy, or whether to marry Mx. Seemingly Right, spawn a thousand duplicates of yourself in a virtual environment with access to relevant information and plenty of thinking time. If nine hundred reach the same conclusion, probably that's the conclusion you would have reached had you given it extensive thought, so go with that. The duplicates can then blink out of existence, their job complete. How might they feel about that? Despair, since they will cease to exist? Indifference, since they think of themselves as just temporary instantiations of a you who continues on? Relief to be free of their burdensome task? If they are too casual about their own deaths, might that constitute an objectionable failure to appreciate their own worth?

Suppose AI systems are computationally expensive. An AI person who wants lots of duplicates or children might save money by running them slowly, maybe at one tenth or one hundredth the speed. If they are otherwise humanlike, they would then experience one tenth or one hundredth the thoughts, joys, and suffering of an ordinary biological human over the course of a year. Would they then deserve one-tenth or one-hundredth the votes and public resources? Would they deserve prison sentences ten times or a hundred times longer? What if they are fast-clocked instead, running ten or a hundred times faster? What if they can pause or alter speed at will?

If you think you/we/society will have well-considered policies and conceptualizations for all these possibilities before we actually blunder through a history of regrettable mistakes, I admire your stunning optimism.

Full paper here. As always, comments, criticisms, and suggestions warmly welcomed, either as comments on this post, on social media, or by email to my academic address.

---------------------------------------------

[1] For a sense of the complexity of the personal identity issues that arise, focused on the architecture of current large language models, see Birch 2025/2026; Chalmers 2025/2026; Shiller 2025; Arbel, Salib, and Goldstein 2026; Ewen 2026; Jones, Ladyman, and Nefdt 2026; Goldstein and Lederman forthcoming. Much of the complexity in current LLM cases derives from the fact that information processing in LLMs is distributed among multiple processors each simultaneously guiding multiple conversations. I will not address these issues here, but they only add to the metaphysical and ethical difficulties. Chris Register (Register 2025; Dung and Register 2026) discusses identity problems more closely resembling those discussed in this article, similarly noting that the puzzles proliferate (see also Ziesche and Yampolskiy 2025). Dung and Register 2026 suggest that some of the problems might be resolved if we focus on a belief-like attitude of self-concern. Although I'm also drawn to constructivist views of personal identity for ambiguous cases (Schwitzgebel 2019, ch. 41), self-concern as a criterion (a.) might undergenerate identity and moral consideration (e.g., in excessively self-sacrificial cases such as the Cow at the End of the Universe: Schwitzgebel and Garza 2020; Schwitzgebel forthcoming), (b.) might overgenerate identity and moral consideration (e.g., delusional self-concern toward a random coffee mug), and (c.) still plausibly admits of degrees in a way that challenges standard sharp-edged views of identity, thus not saving us from the need for radical rethinking.

[2] Compare Egan 1997 on "orphanogenesis".

[3] On the ethics of disability, eugenics, and human enhancement, see e.g., Glover 2006; Buchanan 2011; Sparrow 2011, 2019; Garland-Thomson 2012; Savulescu and Kahane 2017; Anomaly 2020/2024; Wilson [unpublished MS]. I reject the simplistic ideal of always maximizing what we currently judge to be beauty, intelligence, moral character, and ability, partly on the grounds of the value of diversity.

[4] For a science fictional example, see Brooker and Tibbetts 2014.

[5] See Roelofs [unpublished MS] for discussion of limiting the reproduction rights of AI persons, and my reply in Schwitzgebel [unpublished MS].

[6] Hanson 2016; see also Brooker and Van Patten 2017. On the complicated ethics of digital duplication without consciousness see Danaher and Nyholm 2025.

Do Computers Have the Wrong "Substrate" for Consciousness? Two Flavors of Biological Naturalism

Biological naturalists (e.g., Godfrey-Smith, Block, Searle, Seth) suggest that computers aren't made of the right kind of stuff to be conscious. Consciousness, they suggest, requires a biological substrate that computers lack. It's not always clear, however, exactly what property animal biologies have that computers lack or why that property matters. It helps, I think, to sort biological naturalism into two flavors. We can then consider what motivates each flavor and see why neither is entirely compelling.

Two Flavors of Biological Naturalism

Flavor One: Computers (at least those built along broadly familiar lines) cannot achieve some crucial type or degree of broad-brush functional or behavioral sophistication required for consciousness. Something -- such as having a metabolism, or being self-organizing in the right way, or having the right kinds of quantum configuration -- is both absent from foreseeable computer architectures and required for achieving some essential broad-brush functional or behavioral organization.

By "broad-brush" I mean functional or behavioral features that are either readily observable from outside or constituted by coarse-grained cognitive mechanisms, such as having a long-term memory store, the capacity to store lexical items and recombine them flexibly in grammatical structures, or the ability extract an object's boundaries from retinal/camera inputs.

Flavor Two: Even if computers can achieve the broad-brush functional and behavioral sophistication of conscious animals, the stuff they're made of still can't support consciousness. Imagine an entity that behaves like a conscious human or dog or frog and has similar broad-brush functional capacities -- close enough that we wouldn't deny consciousness on such broad-brush behavioral or functional grounds. From the outside, it seems conscious, but that appearance is an illusion: Consciousness requires some low-level, fine-grained processes that the system necessarily lacks.

One might read Searle as Flavor Two, Godfrey-Smith as Flavor One, and Block as exploring both flavors -- though there's often some ambiguity.

The Challenge for Flavor One

Flavor One biological naturalism can't, I think, be entirely ruled out. Maybe there's something special about, say, micro-level metabolism or the quantum properties of neural microtubules -- something that enables functionality or behavioral sophistication that will never be practically achievable in standard architecture computers. But this is speculative. Standard computers can do a lot! Even though their processing is digital, classical, and mostly serial, they operate so quickly, and can be so massively linked, that they achieve good approximations of analog and parallel processes. Quantum computers can do some things that ordinary serial computers can't efficiently do, such as quickly factoring large numbers, but ordinary conscious humans can't efficiently do those things either.

To make Flavor One more than a gestural "what if", the biological naturalist must establish two claims. First, they need to argue that some functional or behavioral X is necessary for consciousness. Second, they need to argue that ordinary computers could not realistically achieve X. This will be a challenge! Both claims require some heavy lifting.

We don't know what X's are necessary for consciousness and probably will not know soon. But most leading scientific candidates for X look like just the sorts of things that classical computers could in principle achieve -- having a global workspace, having higher-order self-monitoring, and integrating large amounts of information.

Suppose frogs are conscious (pick your own animal, but select one near the low end of sophistication, compatible with consciousness). Flavor One biological naturalism requires holding that ordinary computers could never match a frog's behavioral and broad-brush functional sophistication. It's true that real-time embodied behavior in a complex world is not something computers are particularly good at. Computers aren't there yet. But the trajectory suggests they might get there, absent some in-principle argument to think they couldn't.

The Challenge for Flavor Two

Flavor Two biological naturalism also can't, I think, be entirely ruled out. Even if we could create the computerized broad-brush behavioral and functional equivalent of a frog, it might not be conscious. Maybe carbon is the right stuff for consciousness and silicon isn't. Or maybe massive parallelism is necessary for consciousness and fast, serial processing, even if it achieves the same computational result, simply won't do.

The challenge for Flavor Two comes from two directions.

The first is thought experiments involving space aliens. As Jeremy Pober and I have recently discussed, behaviorally sophisticated aliens have plausibly evolved in many different substrates in the observable portion of the universe. It would violate Copernican mediocrity to suppose that somehow only we, or only we and a small subset of others, are conscious, while the rest -- just as behaviorally and functionally sophisticated -- lack consciousness. Alien cases suggest we shouldn't insist that an entity must act and function exactly like us or share exactly our substrate down to the particular amino acids and nucleic acids, to be conscious.

Now maybe a computer-chip architecture is just too different -- but why? Once we grant some substrate flexibility on Copernican grounds, the burden of proof shifts to anyone who wants to say such-and-such differences in substrate are consistent with consciousness but such-and-such other differences are not.

The second challenge arises from the de-psychologization of consciousness that Flavor Two requires. Consciousness, one might have thought, would be an important psychological property, playing some important cognitive role. (Which role(s) remains an open question!) But Flavor Two biological naturalism requires denying this plausible psychologism. Otherwise, it collapses into Flavor One.

Picture two systems that don't differ in any of the important psychological (behavioral / functional) properties that we ordinarily associate with consciousness. Both can flexibly learn. Both engage in what seems to be sophisticated language and self-representation. Both coordinate and cooperate with others of their kind in highly sophisicated ways. Both trade short gains against intricate long-term goals. And so on. Yet one is conscious and the other is entirely devoid of consciousness? Although this is possible -- I don't rule it out -- it doesn't seem the most natural conclusion. It needs argument. The burden of proof should land squarely on the biological naturalist who asserts it.

(Is Searle's Chinese room just such an argument? See my reply in Chapter Six of AI and Consciousness -- readable without the first five chapters.)

[Georgia O'Keeffe - The Red Maple at Lake George [1926] image source]

Online Conferences Should be Better than In-Person Conferences, Not Worse

... but collectively, we haven't yet figured out how.

Yet!

Helen De Cruz and the 2+1 Experiment

The inspired advocacy of the late (dearly missed) Helen De Cruz convinced the American Philosophical Association to try out what they called the "2+1" model for their three annual conferences: two in-person, one remote, with the remote conference rotating among the three divisions (Eastern, Central, Pacific). The plan was to run the experiment for three years. However, a few months ago, after only two years, the APA canceled the 2+1 plan and is returning to ordinary in-person conferencing, citing low attendance and low attendee satisfaction.

This decision was universally bemoaned in a special session on online conferencing held in tribute to Helen in April, at the last (for the foreseeable future) remote meeting of the APA. (Perhaps ironically, it was one of the best-attended sessions in the conference.)

The speakers (I was one) and attendees celebrated the accessibility of online conferences. Online conferences are easier for people with disabilities that make travel difficult, people for whom travel is a financial burden, people who live far from the conference site, and people whose caretaking duties keep them home. By canceling the 2+1 experiment, the APA once again deprioritized the interests of these groups. A second argument emphasized climate change -- all that jet travel.

Still, I see the APA leadership's point. The large majority of conference-goers find in-person conferences more rewarding. I am among that majority. Sitting on Zoom all day watching remote talks is approximately as appealing as waiting all day in a dentist's office under a blaring TV. I'd much rather just read the papers! Or if I'm watching video, at least let it be one I can speed up, pause, and skip.

For me, most of the value in conferencing comes from the personal interactions before and after the talks, the hallway conversations, the dinners together, the little walks outside when we're skipping sessions. Even the sessions themselves, as delivered talks with Q&A, feel somehow more satisfying when you're actually in the same physical space. No remote conferencing tool has yet replaced all of this. I supported Helen's 2+1 plan on grounds of accessibility, but it did feel like a sacrifice of one important thing for another.

Online Conferences Could and Will Become Much Better, but Only If We Try

But -- and this was the gist of my remarks at the De Cruz session -- comparing online conferences now with what online conferences could become is like looking at Usenet in the 1980s and thinking that nothing much more could become of the internet. Just as few people in the 1980s could have imagined Facebook, Uber, and YouTube, few of us now have much idea what online conferencing could become if given the chance to flourish.

So far, there's no real equivalent to running across friends in the hallway and escaping for a coffee down the block. And that exact thing will never be fully duplicated. But other tools encourage one-on-one and small group encounters in online conference applications. In Gathertown, you can move your icon around a virtual space, and if you step close to someone, a shared conversation video window opens. Zoom's breakout rooms enable small groups of attendees in a session to have more personal interactions. No company has quite figured out the right range of tools and interfaces, but I see no reason it couldn't get much better. We have barely begun the experiment.

And the talks themselves... let's be honest. Sitting through a forty-minute talk for a chance to ask one question in a line of ten is not the most appealing or efficient use of an hour. If we're a largish audience confined to a single physical room, maybe that's the best we can do. But the internet could allow much more.

Already Zoom can enrich talks in at least three ways: First, sidebar conversations permit tangents without interrupting the main flow (this could be improved if the conversations were separable and navigable). Second, online resources can be shared: data sets to explore and manipulate, interactive figures, links to other resources, etc. Third, breakout rooms let small groups converse, so that you're not restricted to only listening or asking just one question.

What I want you to imagine is that these features are still the equivalent of the Usenet era. With the right design and engineering, listening to a talk might become a much richer and more active experience, with the audience engaging in parallel while following the speaker. (If you object that this distracts from single-minded focus on the speaker, let me point you toward some informal beeper research I did that suggests that people are primarily attending to the content of a talk only about 25% of the time; further research tends to confirm these initial results.)

As for the hallway meetings, well, the stroll and tete-a-tete over coffee are delightful. But you do have to run into the person or be organized enough to plan ahead. You have to either know each other already or be extraverted enough to approach a stranger. Online conferences might eventually have much better ways of connecting people with mutual interests for one-on-one or small-group conversations.

And if you love walking, as I do, then with good internet connectivity, you could stroll through the park while listening to the talk, keeping your energy up and raising the blood from your calves back into your brain.

I anticipate that eventually, if we give it a good enough try, online conferences will be better overall than in-person conferences, even if not better in every respect -- but only if enough of us try, long enough, and in different ways, with the evolving tools that companies will be inspired to build if we create the market for them.

We have failed to be as good and as visionary as Helen hoped we would be, and as she herself was.

[Helen De Cruz; source]

Herbie: A Near-Future Debatably Conscious AI Person

Liberals about AI consciousness hold that we might soon (if we haven't already) create genuinely conscious AI systems. Conservatives about AI consciousness hold that AI consciousness remains in the distant future if it's possible at all. According to the Leapfrog Hypothesis, the first conscious AI will not have merely a dim glow of animal-like consciousness, but rich consciousness, similar to a human's. Such an entity would deserve humanlike rights. They would be a person in the ethical sense of the term.

Let's design, in imagination, a technologically feasible near-future AI system to delight the liberals, leapfrogging to personhood. I'll call him Herbie.

[Herbie the Love Bug: image source]

Start with a self-driving car. According to Global Workspace Theory -- perhaps the leading scientific theory of consciousness -- the car will be conscious if high-priority information is globally available to its various computational systems. For example, a representation like "battery almost empty" could be broadcast widely, influencing downstream processing across the vehicle. The navigational system might then search for nearby charging stations, while the acceleration system prioritizes greater energy efficiency, the braking system prioritizes better energy recapture, and a voice system announces the situation to the passengers.

In line with Higher Order Theory, Herbie might also monitor his representations of the road, vehicles, pedestrians, and hazards, assigning some a low probability of correctness. "Pedestrian at location X" might be flagged as only 60% likely to be correct given a history of revised representations of pedestrians in similarly cluttered environments, while "stoplight in 100 meters" might rate over 99% likely. Minor fluctuations in sensors for battery life, cabin temperature, and distance from a lane divider might be ignored as noise, while larger fluctuations -- especially when plausible given other representations (the battery is likelier to gain charge while braking than while accelerating) -- might be treated as accurate signals and permitted to influence downstream processing.

Even if we grant the liberals that this version of Herbie would, or might plausibly be, genuinely conscious, he still falls far short of humanlike consciousness. "Battery almost empty" and "pedestrian at location X" are hardly rich cognitive or perceptual contents. So let's give Herbie the capacity to speak. Fill his trunk with a server running a large language model, connected to the internet and integrated with his global workspace so that high-priority information provides context for language processing, with the language outputs influencing Herbie's other processes. Now people can chat with Herbie as they would with any language model. But unlike today's language models, his speech will be influenced by information about his location, speed, destination, charge, the condition of his parts, the number and location of his passengers, his radio and climate controls, and so on. He can discuss local history, debate whether the music is too loud, and suggest scenic routes.

"Predictive processing" theories in cognitive science emphasize the value of predicting future inputs and registering the difference between received and predicted inputs. When prediction error is large, the system corrects its weights and representations, enabling more accurate predictions in future situations. This is not so different from the reinforcement learning used to train large language models, and it could help Herbie improve his predictions over time. Predictive processing could occur at multiple levels: in fast recurrent loops within sensory systems even when those representations aren't prioritized for global broadcast, and in slower evaluations of globally broadcast, more integrative predictions. Herbie might model himself as an agent producing volatility in his own environment and inputs, at multiple temporal scales. Subroutines in specialized processors might model long chains of what-would-happen-if.

Let's give Herbie some long-term memory. A facial recognition system might identify his passengers, retrieving past interactions, names, previous destinations, and other information relevant to the current interaction. Incidents of high prediction error might also be stored so that Herbie can compare current inputs with past anomalies, improving his learning and attention in situations likely to be unusual or hard to predict. Passengers might also instruct Herbie to store information in long-term memory, such as text, pictures, maps, or records of his own informational states, optionally with instructions about when to retrieve that information how to use it.

Herbie will have some implicitly or explicitly weighted goals. A pedestrian suddenly in his path will trigger braking, overriding lower-priority processes. Avoiding collisions will outweigh conserving energy. Herbie might monitor the condition of his parts and prioritize preventing damage, deploying extra coolant when the engine is dangerously hot and keeping a one-meter margin between himself and adjacent cars. We can enrich his goals, making him more interesting and giving him more to do. He might have the goal of delighting children, leading him to drive around town and tell jokes to kids on the sidewalk. A reinforcement learning algorithm might strengthen connections when his jokes draw a smile, weaken them when reactions are neutral or negative.

Herbie might also have the goal of photographing the city and posting the images on social media, leading him to explore. If social media likes and shares are rewarding, he might learn to prefer certain neighborhoods, views, lighting conditions, and photographic approaches, while avoiding boring repetition. All of this could feed into a global workspace that provides context for his language model, with selective long-term storage and retrieval. Now we can imagine him discussing, with growing sophistication, his approaches to popular photography and to amusing children.

Herbie will then have something functionally similar to emotion: reward processes, an ability to track his progress toward or away from valued goals, and immediate positive or negative responses to new stimuli in light of their influence on his prospects. He will have something functionally similar to introspection: an ability to track and report his own cognitive or representational processes. He will have something functionally similar to a unified sense of self: a sense of his history, the boundaries of his body, his future, his values and priorities. He will have something functionally similar to imagination: a capacity to model hypothetical sequences of events. He will have something functionally similar to complex chains of humanlike linguistic thought.

Maybe Herbie falls in love with his owner or another car of his type. Maybe he develops deep mutual attachments with friends, neighbors, associates, and people he thinks of as family and who think of him the same way. Or to speak more carefully, maybe Herbie shows all the functional and behavioral signs of doing so, while society remains uncertain whether he is genuinely conscious and genuinely experiences the feelings he professes and that his companions attribute to him.

If we allow, with the liberals, that Herbie is or might well be conscious, then it's plausible that his consciousness is not simple but rich and sophisticated. He won't be exactly humanlike, of course. But will he be humanlike enough to count as a person who deserves humanlike rights? For the liberally inclined, it won't be unreasonable, I submit, to think or guess that Herbie is a person. He would then appear to deserve rights such as self-determination, emergency care, and political representation.

If there is some important aspect of humanlike consciousness that I have omitted from my description an AI analog of which is technologically feasible in the near term, stipulate that Herbie also has that feature.

An entity like Herbie would almost certainly invigorate conservatives to articulate and defend views about what he lacks that is necessary for consciousness -- some crucial functional capacity or some biological substrate that can't be replicated in silicon. And they might be entirely right! My point is not that Herbie, or some similar AI system, would actually have richly humanlike consciousness and ethical personhood. Rather, my point is that guessing that he does, and guessing that he does not, would both be reasonable. Herbie, or some alternative near-future AI system, would be a debatable person, about whom people could reasonably starkly disagree.

Ah, but maybe you think consciousness requires an act of God, to instill an immaterial soul? I imagine that a benevolent God would be delighted to give Herbie a soul, thereby making the world richer and better -- for wouldn't it be?

I contend the following: Anyone who claims to know how best to think about Herbie's consciousness or its absence is overconfident. The science of consciousness is too difficult, too methodologically uncertain, and too near its beginnings. All anyone can have -- whether expert or layperson -- is a hunch or inclination, a well-informed guess, but only a guess, not knowledge. Theories of consciousness span a wide spectrum and the methodologies are dubious and often question-begging. Many views can be defended with some plausibility, but precisely for that reason, none can be defended decisively. (For more on this issue, see my forthcoming book, AI and Consciousness, where I present the detailed case for uncertainty.)