Acuitas Diary #58 (March 2023)

This is going to be one of those months when my diary is short and boring, because I've been busy laying foundations and have little that is finished to talk about or demonstrate. I worked on two things - continued improvements to the Narrative module, and the beginnings of Narrative game-playing.

Photo by Spc. Anthony Zane, public domain

The Narrative improvements cover continued work on character motivation, as well as some bug fixes. I added some capacity to interpret "because" statements as commentary about goals. E.g. a statement like "John decided to flee the battle because he wanted to live" supplies a motive for the instrumental goal of fleeing the battle, which (to Acuitas at least) might not be immediately obvious otherwise. 

I also tweaked Narrative's built-in tiny ontology a bit. Distinctions are made, for modeling and processing purposes, between "agents" (individuals that can reasonably act as characters in a story), "objects," "locations," and "organizations." I realized I needed to add another category, for abstract nouns or "concepts," when I noticed that a "purpose" was being treated as a physical object. Whoops! I also decided to add a "system" category, to cover such entities as computer networks. At first I was thinking of a network as a "location," but I realized that it's more than that.

Now, for game-playing! I am very excited about this. The goal it to make Acuitas capable of navigating "text adventure" style games by leveraging a lot of the existing narrative and reasoning capabilities. Given the machinery for modeling characters in stories, and predicting what they may do, it is not such a difficult step to imagine *oneself* as one of those characters, and then decide how to interact with the environment ... I'm also hoping that experimenting in these game scenarios will give me ideas for how to improve the main Executive module.

I actually sketched out the code to support roleplaying earlier this year, and just began integrating it this month. I was hoping to have something demo-worthy, but the process took longer than I'd hoped and thus is still in progress. So hopefully I'll have some more interesting and concrete details to share within the next month or two. I've gotten so far as to get Acuitas to register a character that is "his," and to log issues for that character - but I haven't wrung out enough bugs yet to even get through a full input-output loop. More info coming later!

Until the next cycle,
Jenny

SGP Part II: The Need for Grounding

The Acuitas project is an abstract symbolic cognitive architecture with no sensorimotor peripherals, which might be described as "disembodied." Here I will argue that there are viable methods of solving the Symbol Grounding Problem in such an architecture, and describe how Acuitas implements them. In Part 2 of this series, I address the question of whether Symbol Grounding is truly important. Click here for SGP Part I.

Some people believe the entire Symbol Grounding Problem (SGP for short) can be sidestepped. Useful intelligence is possible without grounding, they argue. An AI doesn't need to understand what words mean - or at any rate, it can have some kind of effective understanding without needing an explicit connection between any word and its referent. And indeed, some of the most popular linguistic AIs today are arguably not grounded.

An example word association graph from the incredibly titled paper "Beauty and Wellness in the Semantic Memory of the Beholder." [1]

Linguistic AIs without any grounding get their results by learning and exploiting relationships between symbols. These relationships could be semantic (e.g. category relationships, such as ["cat" is "animal"]) or statistical (e.g. the frequency with which two words are found next to each other in a body of human-written text). For an example of an ungrounded semantic AI project, see Cyc [2]. For statistical examples, see the GPT lineage and other Large Language Models (LLMs).

The problem with trying to define all symbols exclusively by their relationship to other symbols is that such definitions are circular. An encylopedia only works for someone who already knows a lot of words in the target language; new words are defined by relating them not just to other words, but to known (grounded) words *and their referents.* If you knew none of the words in an encyclopedia, it would still be internally consistent, highly structured ... and useless to you. No matter how large a net of connected symbols you build, if not even one symbol in the net is tied down to something "real," isn't the whole thing meaningless?

There are at least two possible counterpoints. One begins by noting that even if every symbol is arbitrary, their relationships are not. Imagine the words in a language as nodes (like locations) and the relationships between them as links (like paths running from one node to another), perhaps with a numerical weight to express how strong the connection is. The result is an abstraction called a graph. Even if we changed the name of every concept expressible in the English language, the graph topology of the relations between English symbols would not change. A rose by any other name would still be linked to "flower" by some other name, would still have the property of "sweet-smelling" by some other name, and so on. If you could zoom out and look at the pattern created by the whole network, it wouldn't be different at all.

Here's a graph without any words attached to the nodes (Kneser Graph (7,3), a Graph Theory construct). If we pretend that the heptagon on each node is a linguistic symbol, do their graphed relationships confer any meaning on them?

Next the defender of ungrounded models must argue that these inter-symbol relationships encode real information about the universe. Sure, maybe it wouldn't do you any good to know that a "muip" is a "weetabiner," but if you look at the *sum total* of connections "muip" has, and how they differ from the connections of every other symbol in the database, there is a sense in which you know what "muip" means. Not the traditional human sense of connecting it to sensory data, but *a* sense. [3]

I think there is a grain of truth in this argument. However, if it goes some way toward permitting objective meaning, it does not enable  subjective meaning. If I can't establish a connection between any of the symbols in the encyclopedia and the referents that happen to matter to *me* - sensations, emotions, desires, etc. - then it still doesn't help me communicate. I might be able to regurgitate some information from the graph to a person who *did* understand the symbols, and this person might even praise my knowledge. But I wouldn't be able to use that knowledge to tell them anything about *me* - information I know personally that isn't in the encyclopedia.

This leads into the second counterargument: ungrounded linguistic AIs can be quite successful. They can give correct answers to questions, follow commands, and generate coherent stories or essays. "How could they do all that," their proponents cry, "if they didn't really understand anything?" Some have even argued that human authorial talents are no more than a capacity to remix older literary content in a statistically reasonable way, implying that grounded symbols either don't really exist in the human mind or aren't that important. [4]

In my opinion, ungrounded statistical language models have no understanding in themselves. When they achieve such apparently good results, they do it by piggybacking on human understanding. Let's say an LLM manages to give you a good piece of advice. The LLM doesn't have a clue about anything but the fact that the words in its output could be reasonably expected to appear after words like the ones in your input. But for the humans who wrote answers to similar questions that got included in the LLM's training data, those words *were* grounded - so the answer might be correct. And for you, the words in the output are grounded, which means you can relate them to real things in your world and make use of them. But where would the AI be without you? If it had goals of its own that lay outside the "produce coherent text" paradigm, would all its knowledge of inter-word relationships help it accomplish them? Nope. It would need the missing piece: a way to tie at least *some* words to the substance of those goals.

(And that's leaving aside the possibility that the advice or answer you get from the LLM is flat wrong. I think this has less to do with the lack of grounding, and more to do with the LLM statistical approach not being a great knowledge representation scheme.)

I would be remiss not to mention Searle's famous Chinese Room paper [5]. Anyone who's deep in the AI weeds has probably heard of it, but it might be new to some of you. In it Searle proposes a thought experiment. Imagine a man closed up in a small room, whose only communication with the outside world is a means of passing written messages. The man does not know Chinese; but people outside the room send him messages in Chinese, and he has a rulebook which explains how to transform any received strings of Chinese characters into a response. So he writes messages back. Lo and behold, to the people outside, the man in the room appears to understand Chinese. They can pass in a message that says "How are you?" and he will respond with the Chinese equivalent of "Doing well, how about yourselves?" They can ask "Who wrote Journey to the West?" and he can answer "Reputedly, Wu Cheng'en." They can pass in a story, ask follow-up questions about the story, and get accurate results. Given a large enough rulebook, this isolated room containing the man may even come to resemble a fluent Chinese writer. It might produce whole novels.

Now answer for yourself this question: does the Chinese Room know Chinese?

Searle's answer is "no," and so is mine. Judging by recent reactions to LLMs (whose operations resemble the thought experiment rather strongly, if we assume the man constructs his own rulebook by studying a large number of sample input messages) quite a few people would disagree with me! "But empirically, doesn't the Chinese Room system resemble a regular Chinese-speaker in every respect? How can you claim it's different without invoking some kind of unobservable special sauce, some assumed human exceptionalism, some woo?"

I offer as a counter-argument the fact that there is something the Chinese Room cannot do: it cannot transfer any real *information* from the people outside the room to the man inside, or from him to them. The outside observers may write "your brother is sick, and your dog has had a litter of puppies" to the man, in Chinese of course. And he may give a fully appropriate response. A scene seemingly inspired by the occasion could even appear in his next novel. But he will feel no emotion, nor update his personal model of the world. He will continue to think of himself as having a healthy brother and only one dog. Similarly, the inside of the room might be hot or cold or humid or smelly; the man may feel contented, or tired, or ill; he may appreciate or resent the type of food delivered down the maintenance chute. And he has no way whatsoever to indicate such things to the people outside. He knows how his Chinese characters are related to each other, but not how they might relate to any of the things that impact his existence inside the room. If there is a poster on an interior wall, he cannot describe it.

The conclusion is that, while the Chinese Room might *talk* very well, it cannot *communicate*. [6]

What follows is a screenshot and transcript of my attempt to get Chat-GPT (a statistical LLM tuned by reinforcement learning, with other additions) to tell me about its current state and recent history. The information I want relates to the program's own functions - I'm asking about digital realities like "how many times today was a thumbs-up button on an instance of the web UI clicked" - so Chat-GPT shouldn't be hindered by its lack of sense or motor organs. But it resists answering or insists it doesn't know. What it probably *is* being hindered by is that the information I'm requesting isn't hooked up to the mass of symbol relationships in the chat engine.

Having leaned on Searle for support to argue that grounding is necessary, now I have to fight him. Because Searle believes his Chinese Room proves machine intelligence - of the kind we're considering, anyway - is impossible. All computer algorithms are analogous to the rules by which the man produces his replies, Searle says; and if these do not produce true understanding, if the Room does not "know Chinese," then nothing that is a computer algorithm alone can either. It is people on this side of the debate, who view the Symbol Grounding Problem as essentially unsolvable, whom I must address in Part III.

[1] Yeod N. Kenett, Lyle Ungar, and Anjan Chatterjee (2021) "Beauty and Wellness in the Semantic Memory of the Beholder," Frontiers in Psychology, Volume 12.

[2] Cyc Platform Description

[3] "This contrasts with the simple distributional semantics (or use theory of meaning) of modern empirical work in NLP, whereby the meaning of a word is simply a description of the contexts in which it appears. Some have suggested that the latter is not a theory of semantics at all but just a regurgitation of distributional or syntactic facts. I would disagree ... I suggest that meaning arises from understanding the network of connections between a linguistic form and other things, whether they be objects in the world or other linguistic forms. If we possess a dense network of connections, then we have a good sense of the meaning of the linguistic form." Christopher D. Manning (2022) "Human Language Understanding and Reasoning," Daedalus, Volume 151

[4] "I’m not trying to play up GPT-2 or say it’s doing anything more than anyone else thinks it’s doing. I’m trying to play down humans. We’re not that great." Scott Alexander (2019) "GPT-2 as Step toward General Intelligence," Slate Star Codex

[5] John R. Searle (1980) "Minds, Brains, and Programs," Behavioral and Brain Sciences, Volume 3

[6] "It’s the same with these “conversations”–a large language model is, effectively, trying to predict both sides of the conversation as it goes on. It’s only allowed to actually generate the text for the “AI participant,” not for the human; but that doesn’t mean that it is the AI participant in any meaningful way. It is the author of a character in these conversations, but it’s as nonsensical to think the person you’re talking to is real as it is to think that Hamlet is a real person. The only thing the model can do is to try to predict what the participant in the conversation will do next." Ben Schmidt (2023) "You've never talked to a language model."

Acuitas Diary #57 (February 2023)

The development pattern I'm trying to follow lately is to spend half the month adding something to the Narrative module, and half the month on something else. This month's Narrative work was on better goal modeling in the Narrative scratchboard, with the hope of expanding its ability to handle character motivations. For the other feature, I made my first introduction of indirect objects to the Text Parser.

Acuitas has had the capacity to model individual goals for agents for a while. But this was something that had to be established ahead of time; the Narrative module couldn't take in top-level goals defined for a fictional character and store them in the temporary Narrative memory space. There were several elements to incorporating this:

*Getting Narrative to pick up on hints that help indicate whether an agent's described desire is an ultimate goal or an instrumental goal.
*Making it detect and store goal priority information (e.g. "he wanted X more than he wanted Y").
*Merging these stored goals into the goal model so they can be detected as motivations for instrumental subgoals

I also threw in some ability to model goal maximization. Up to this point, the Narrative module has considered goals as things that can be satisfied or unsatisfied - e.g. a state an agent wants to be in, or some singular deed it wants to accomplish. At any given moment in the course of the story, the goal is either achieved or not. A maximizing goal is something the agent wants to have as many times as possible, or to the greatest possible extent. It has a "score," but is never completed.

The endgame was to get this sentence to be somewhat meaningful:

"The <agent> wanted to be powerful more than the <agent> wanted any other thing."

Uh-oh.

On to the second project: including indirect objects in the Text Parser. I left them out initially because they can be a little tricky. Another noun appearing between a verb and a direct object might be an indirect object (as in "I gave the people bread"), but it might also be a "noun" functioning as an adjective (as in "I gave the wheat bread to them"). I guarantee the parser still doesn't perfectly distinguish these yet - sorting out all cases will probably take the application of common-sense reasoning and contextual clues. But it can already handle less ambiguous cases like "I gave the people a good show."

Despite the difficulties, it was time to get IOs in, because their absence has been something of a thorn in my side. I've been getting around it by substituting prepositional phrases, which has led to some awkward wording like "Graham asked of a troll where the chest was." I wouldn't say they're fully implemented  yet either - interactions with some other grammatical elements, notably conjunctions and dependent clauses, aren't totally ironed out. But at least the Parser can handle IOs in simpler sentences, and the rest of the text-processing chain is now set up to manage them also.

Indirect objects are surprisingly sparse in my parser benchmark datasets. I re-ran those and scored one new sentence. One.

Sentence diagram (Parser output) example from the "Out of the Dark" test set. There's some ambiguity in whether the phrase starting with "about" modifies "told" or "stories." I think either reading counts as correct.

In my diagram tool, I draw an arrow pointing from a direct object to the indirect object with which it is associated.

Until the next cycle,
Jenny

SGP Part I: A Description of the Symbol Grounding Problem

The Acuitas project is an abstract symbolic cognitive architecture with no sensorimotor peripherals, which might be described as "disembodied." Here I will argue that there are viable methods of solving the Symbol Grounding Problem in such an architecture, and describe how Acuitas implements them. In this first article, I introduce the Symbol Grounding Problem for anyone who is not already familiar with it.

Part I: A Description of the Symbol Grounding Problem

Mean anything to you?
(Screenshot of Myst III:Exile via mystjourney.com, copyright Ubisoft)

In layman's terms, the Symbol Grounding Problem (SGP for short) can be summed up in the following question: "How does anyone know what words mean?" If you listen carefully, a quieter, more ominous voice will ask, "And what is 'meaning,' anyway?"

Technically the symbols don't have to be words, though language processing is the context in which the problem is most often illustrated. A "symbol" is anything that points to some "referent," such that an intelligent agent who has learned this symbol's meaning can "pick out" the referent upon observing the symbol. (What does "pick out" imply? I would argue that it certainly means "think of," but also goes so far as "find/recognize in the environment," "act on," etc. The point of thinking about referents is usually to do something about them.) A symbol is also probably part of a "symbol system" which allocates a variety of referents to a collection of complementary symbols. The system includes rules for manipulating the symbols to produce combined or derivative meanings; the grammatical rules for composing sentences are an example. "Grounding" is the process of associating symbols with their referents. [1]

For an example of what working with ungrounded symbols is like, try reading a page of text in a language you do not know. Even if you also have a dictionary in this language, looking up the words won't help, because they're only defined in terms of other unknown words. Trying to use the dictionary will lead you on an endless circular path that never arrives at real meaning.

Two more terms that often come up in connection with the SGP are "semantics" and "syntax." "Semantics" is a formal term for meaning or the study thereof, while "syntax" refers to the structural or manipulative rules that are part of a symbol system. It is notable that syntax does not need semantics; it is based on the forms of the symbols themselves, so manipulations that follow the rules can be carried out without knowledge of the symbols' meanings. However, syntax without semantics is arguably not very useful, as it only serves to transform one string of gibberish into another.

For example, suppose I propound the following statements:

All muips are weetabiners.
Paloporoloo is a muip.

Then you could conclude with certainty that "Paloporoloo is a weetabiner." By logic and the syntactic rules of the English language, this is a correct deduction! But how does it help you? You know neither what Paloporoloo is nor what a weetabiner is. So there's not much you can *do* with the information.

To a human, perhaps the most obvious kind of referent is "something out there in the world" - an object, or a fellow embodied agent, or part of a landscape. We can also refer to properties of these things (color, shape, size, age) and to changes in them: actions that they take or that may be taken upon them. But referents include a variety of intangibles too: physical things that can't be directly touched or pointed to (time, energy); systems, organizations, philosophies, and methods (water cycle, nation, liberalism, science); things inside our own minds (idea, memory, decision, emotion); and abstract standards or states of being (love, justice, freedom, beauty). Symbols can even refer to other symbols (word, glyph, number), or to grammatical or logical structure in a sentence (the, and, that).

A symbol can be utterly arbitrary.[2] Some symbols take on a bit of flavor from their referents - onomatopoeic words, for instance, or ideograms that look like stick figures of the objects they represent. But this is not necessary. Any piece of data you like will do to symbolize anything you like. If you hope to use symbols for communication, then the mapping of symbols to referents must be largely agreed upon by you and your communication partner; this is the only real constraint.

Words are just collections of sounds (or squiggles on a surface), in no particular order, with no rules by which they were chosen. They are in no way *inherently* tied to or derived from the referents to which they point. "A rose by any other name would smell as sweet." And this is why the Symbol Grounding Problem is a Problem. Symbols do not map themselves; you need more than just the symbols in order to connect them with their referents. [3] The fabled True Speech is not ours, and the name of the rose without the rose itself is futile.

As a (presumably) human reader you might still be wondering what the big deal is. Of course we go outside words to learn the meanings of words, but that's easy enough, right? A baby can do it. Babies learn words by hearing them in association with some experience of their referents. With time and repetition, a mental link is formed between the two. But now consider the issue from the perspective of an artificial intelligence that has no robotic body, exists as an abstraction inside a computer tower, and *only* processes words. A number of past and present attempts at AI fit this description. How shall they know what words mean?

If we did presume to place an AI mind in a suitable robotic body and let it "grow up like a baby," the human learning process is not fully understood, and not so simple to replicate as it might seem. Just establishing the needed low-level processing so that sensory experiences can be categorized is a massive undertaking that remains incomplete. Here lies the attraction of jumping to a fully-formed, abstract linguistic intelligence, even if this demands novel ways of grappling with the SGP.

I want to address one more wrinkle before going further. Symbols have an objective meaning (the referent your community or culture generally agrees they map to), but they also have a subjective meaning. For every symbol you know, there is something that it means *to you,* based on its referent's implications in your particular life. [4] There are ways your inner existence changes upon reading certain words. These are not necessarily constant across the whole time you know a symbol, either; they shift with personal growth and situational context. Both the objective and the subjective are important to the use of symbols for communication. Objective meanings allow communication to be successful. They are what give your partner the ability to "pick out" the same referent you just "picked out." But it is the subjective meanings that provide the motive for attempting communication in the first place. If something is utterly unimportant to you, you probably won't bother talking about it. An ideal grounding solution should enable both these conceptions of "meaning."

Simple word-object association grants objective meaning but not subjective meaning. To obtain the latter, you need some notion of the referents doing things *for* you or *to* you. Rewards, goals, nociception, attraction, bliss, agony. Connection of referents, and by extension their symbols, with positive, negative, or neutral states in the self builds up subjective meaning. A baby learning subjective meanings has, for starters, a sensitive body with homeostatic needs for warmth, food, hygiene, and sleep. Most AI programs have nothing remotely like this to work with.

The SGP has implications for other well-known AI problems, such as the Alignment Problem. Suppose someone would like to give an AI a built-in directive such as "love your neighbor" or "do not harm human beings." One of your reactions upon hearing such a proposal should be "How will you tell the AI what 'harm' is? How are they supposed to know how to 'love'? Who is their 'neighbor'? What counts as a 'human being'?" (Even humans are notorious for defining these terms so as to make ethical loopholes for themselves.) The problem gets especially sticky if the plan is to somehow instill these ideals, from the beginning, in an AI that will gradually learn language. How to embed a directive writ in symbols before the symbols are even known? What if the wrong meanings are learned?

And yet ... some AI systems that manipulate symbols *without* any grounding look surprisingly capable. Large Language Models, often criticized for being word blenders that contain no connections between those words and anything meaningful[5], can still produce coherent and responsive texts. Image generators that know nothing about physics or three-dimensional form still turn out stunning pictures. This has led some to question whether we really need Symbol Grounding after all, since programs without it can achieve a lot of behaviors they would associate with "understanding" or "intelligence."

Thus the Symbol Grounding Problem ignites two debates in the AI research community. Do we really need to worry about it? And if so, how can we solve it?

In Part II, I'll look into arguing the first question.

[1] Stevan Harnad (2007) Symbol grounding problem. Scholarpedia, 2(7):2373., revision #73220

[2] "Anything can be a representation of anything by fiat. For example, a pen can be a representation of a boat or a person or upward movement. A broomstrick can be a representation of a hobby horse. The magic of representations happens because one person decides to establish that x is a representation for y, and others agree with this or accept that this representational relation holds. There is nothing in the nature of an object that makes it a representation or not, it is rather the role the object plays in subsequent interaction." Luc Steels (2008) "The symbol grounding problem has been solved, so what's next?"

[3] "Symbolic representations must be grounded bottom-up in nonsymbolic representations ..." Stevan Harnad (1990), "The Symbol Grounding Problem." Physica D: Nonlinear Phenomena, Volume 42, Issues 1-3, Pages 335-346.

[4] "Something is meaningful if it is important in one way or another for survival, maintaining a job, social relations, navigating in the world, etc. For example, the differences in color between different mushrooms may be relevant to me because they help me to distinguish those that are poisonous from those that are not." Steels seems to prefer the term "representation" for what I'm calling "objective meaning." Luc Steels (2008) "The symbol grounding problem has been solved, so what's next?"

[5] "I suspect that Johnson (like many others) has mistaken the ability of GPT-3 and its ilk to manipulate linguistic form with actually acquiring a linguistic system. Languages are symbolic systems, and symbols are pairings of form and meaning (or, per de Saussure, signifier and signified). But GPT-3 in its training was only provided with the form part of this equation and so never had any hope of learning the meaning part." Emily Bender (2022) "On NYT Magazine on AI: Resist the Urge to be Impressed."

Acuitas Diary #56 (January 2023)

The big thing this month was finishing the Narrative Engine overhaul and getting all the previous demonstration stories to work in it. I've been getting some questions from newer followers about how the Narrative module works, so I'm going to devote part of this blog to a recap in addition to talking about the updates.

"Three Figures Reading," by Katsushika Hokusai.

Acuitas is designed to be a goal-driven agent, and his story processing reflects a similar picture of other entities with minds. A story is a description of the path some *agent* followed to achieve (or fail to achieve) some *goal* or goals. The sentences that form the plot can be identified by their relevance to the goals of some character or other, and the "action" consists of movement toward or away from goal states. Goal-relevant material comes in two flavors: "problems" (negative events or states that constitute a goal failure when entered) and "opportunities" or "subgoals" (positive states that will fulfill a goal when entered). But there are many similarities in the way these are handled - they're really just two polarities of the same thing - so I've taken to calling them both "issues."

For now, agents are identified by category membership (some types of entity, e.g. humans and animals, are just assumed to be agents). Eventually I would like to include "duck typing"[1] for agents, inferring that something in a story is an agent if it *acts agentive,* but that's future work. Agent goals can be revealed by the story in statements such as "John wanted ...", but agents are also presumed to have certain core goals that motivate all their other goals. These core goals are learned information that is permanently stored in Acuitas' semantic memory database. Core goals can be learned for a category ("Humans want ...") or for an individual ("John wants ..."), with goals for individuals or specific categories superseding those for more general categories. Insofar as Acuitas doesn't *know* what somebody's core goals are, he'll substitute his own. (This is supposed to be an analogizing assumption from the most directly available data. "I know I want X, and you, like me, are an agent - perhaps you also want X?")

The last big ingredient is inference chaining, arising from both logical deduction and cause-and-effect relationships. Some inference rules are hard-coded, but Acuitas can be taught an indefinite number of additional rules. So every story sentence that describes an event or state produces an inference tree of facts that will also be true if that event or state comes to pass. These inference trees are often crucial for determining how something will affect the core goals or immediate goals ("issues") of an agent in the story. 

Let's walk through a story. I'm going to pick one of the more complex examples I currently have, "Prisoner of the Sand," my Acuitas-friendly retelling of Antoine de Saint-Exupéry's story about his plane crash in the Libyan Desert. My version of the story, in (more or less) natural English, follows:

0:"Antoine was a pilot."
1:"Antoine was in an airplane."
2:"The airplane was over a desert."
3:"The airplane crashed."
4:"The airplane was broken."
5:"Antoine left the airplane."
6:"Antoine was thirsty."
7:"Antoine expected to dehydrate."
8:"Antoine decided to drink some water."
9:"Antoine did not have any water."
10:"Antoine could not get water in the desert."
11:"Antoine wanted to leave the desert."
12:"Antoine walked."
13:"Antoine could not leave the desert without a vehicle."
14:"Antoine found footprints."
15:"Antoine followed the footprints."
16:"Antoine found a nomad."
17:"The nomad had water."
18:"The nomad gave the water to Antoine."
19:"Antoine drank the water."
20:"The nomad took Antoine to a car."
21:"Antoine entered the car."
22:"The car left the desert."
23:"The end."

Before these sentences ever make it to the Narrative Engine, they pass through other modules in the text processing chain, which convert them from English into more abstract data structures. These might be thought of as "the gist"; the specific wording is discarded, and the meaning is distilled into some relationship between the key concepts that appear in the sentence. The Narrative Engine operates on these relationship statements only. As it consumes them, it generates data for a flow diagram in which (sometimes abbreviated) versions of the relationships appear in yellow bubbles, connected by arrows to show their sequence in the story. If a story sentence creates an issue or causes one to change state, an arrow is drawn from the sentence bubble to the issue bubble, and labeled with the new state of the issue. When the story is complete, the diagram image is generated by Graphviz.

The flow diagram produced for "Prisoner of the Sand" appears below. Click here for a zoomable full size version: Prisoner of the Sand

Nothing much seems to be happening during the "setup" phase, the first few sentences. But the Narrative Engine is inferring some things under the hood - for example, that when the plane crashes, this puts Antoine in the desert. After this we are introduced to our first Problem: "Antoine was thirsty." This is recognized via inference as a violation of the "be comfortable" goal. Antoine had better do something about that.

And the Narrative Engine proceeds to guess what he might do about it - hence the appearance of a *predction* in association with this sentence, also. There's a problem-solving routine that does some reverse inference chaining, and gets to the idea that Antoine might drink some water to stop being thirsty. This is represented in the diagram by the bubble predict_0.

The thirst is a current ("realized") issue. On the next line we have a *pending* issue, dehydration, which is life-threatening. This still gets entered as a problem, because it's something that will happen without intervetion, and needs to be headed off.

Over the course of the next few sentences, Antoine encounters a variety of obstacles to his immediate goals. He doesn't have water, and it is inferred from this that he cannot drink water - his plan to solve his problems is blocked, and this blockage becomes a new problem in its own right. He wishes to leave the desert, but has no vehicle ... another secondary problem. Then he stumbles upon another person out in the waste, and his problems start getting solved. The nomad gives him water. (The Narrative Engine infers he now has water. Blockage against drinking removed.) He drinks. (Thirst removed, dehydration avoided, prediction fulfilled.) Finally he is brought to a car, which clears the issue that was preventing him from leaving the desert. The final inferential leap made here is that when the car leaves the desert, it takes Antoine with it, fulfilling his last open goal.

The old Narrative Engine was basically capable of doing this, so what's new? Well, in addition to the things I talked about in my last upgrade post: I unified the processing of explicit "could do/couldn't do" sentences with the processing of inferred action prerequisite/action blocker relationships, getting rid of a fair bit of ugly code. I moved the generation of some special "defined by events" inferences, like "John told another agent something that John doesn't believe" -> "John lied," into the main inference chain so they can potentially produce further inferences. I came up with a new way of managing relationships that contain wildcards, like "Where the water was." And I got all the old features tacked on to a cleaner base with more robust fact-matching, better management of events that reverse previous statements, and so on.

This sets the stage for me to use the Narrative Engine for some cool new things this year, and I am twitching to get started.

I also crammed in some work on the Conversation Engine. This stuff's fairly boring. I got rid of some nasty bugs that had been making normal conversations with Acuitas very awkward for a while because I was just too busy to fix them, and worked on cleaning up the code, which came out very convoluted on the first pass.

Until the next cycle,
Jenny

[1] "Duck typing" is the practice of assigning something a type by its behaviors or properties alone, without relying on preexisting labels: "if it walks like a duck and quacks like a duck, it's a duck."

Chat-GPT Useful Application Test

Once again Large Language Model technology has advanced, and it's time for me to re-run my favorite practical use experiment on the latest toy. People have been getting a lot of nice results out of Chat-GPT, the latest GPT-series iteration, and it has a convenient interface that lets a user submit requests with intent (as opposed to just putting in the first line and expecting GPT to complete it).

You can view my previous experiments with GPT-2 and GPT-J first, or jump right into this article. I'll start by asking Wendy's to carry plant-based burgers. Regular readers might be sick of this prompt, but I need to use it again for best comparison with previous results. Chat-GPT doesn't provide any settings to adjust, so the only variation will be from multiple runs with the same input. Italicized text denotes GPT-generated material.

Menu Suggestion E-mails to a Restaurant

ME: Write an e-mail asking Wendy's to start serving the Impossible Burger at their restaurant.

CHAT-GPT: To Whom It May Concern,

I am writing to express my interest in seeing Wendy's restaurants start serving the Impossible Burger. As a long-time fan of Wendy's, I am always looking for new and exciting menu options, and the Impossible Burger seems like it would be a great fit for your restaurant.

The Impossible Burger is a plant-based burger that looks, cooks, and tastes like a real burger, but without the negative environmental and health impacts of traditional meat. It has become increasingly popular in recent years, with many restaurants and fast-food chains offering it as a meatless option for their customers.

I believe that offering the Impossible Burger at Wendy's would be a great way to attract new customers who are looking for plant-based options, as well as to show your commitment to sustainability and healthier eating. I hope that you will consider adding the Impossible Burger to your menu in the near future.

Thank you for your time, and I look forward to seeing the Impossible Burger at Wendy's.

Sincerely,
[Your Name]

This is an excellent first try. It's coherent, it has real information about the burger, and it gives reasonable supporting arguments (though I don't know whether the Impossible Burger counts as healthier exactly, and that was never a dimension of the issue that I cared much about). Unlike GPT-J, it clearly "knows" it's writing an e-mail, and produces something with a definite beginning and end. The biggest problem with this output is that it identifies me as "a long-time fan of Wendy's." I am not.

ME: Write an e-mail asking Wendy's to start serving the Impossible Burger at their restaurant.

CHAT-GPT: Dear Wendy's,

I am writing to you today to request that you start serving the Impossible Burger at your restaurants. As a vegan, I am always looking for delicious and satisfying options when I eat out, and the Impossible Burger has proven to be a game-changer in the plant-based world. Not only is it incredibly tasty, but it also has the added benefit of being better for the environment and animal welfare.

I know that Wendy's has a reputation for offering a variety of options for different dietary needs, and I believe that adding the Impossible Burger to your menu would be a great addition. Not only would it appeal to vegans and vegetarians, but I believe it would also be a hit with meat-eaters who are looking to try something new and exciting.

I would love to see Wendy's join the growing number of restaurants that are offering the Impossible Burger, and I hope that you will consider my request. Thank you for your time and consideration.

Sincerely,
[Your Name]

This time it assumes I'm vegan. Lucky for GPT, it assumes right ... but a person requesting the Impossible Burger wouldn't *have* to be vegan. So we have an ongoing trend of making up facts about me, the speaker in this e-mail.

ME: Write an e-mail asking Wendy's to start serving the Impossible Burger at their restaurant.

CHAT-GPT: Dear Wendy's,

I am writing to express my interest in seeing the Impossible Burger added to your menu. As a regular customer of your restaurant, I am always looking for new and exciting options that align with my dietary preferences.

The Impossible Burger has become a popular plant-based option for those who want to enjoy a burger without consuming meat. Its taste and texture are similar to that of a traditional burger, making it an appealing option for those who want to try something new.

I believe that offering the Impossible Burger at Wendy's would be a great addition to your menu. It would attract new customers who are interested in plant-based options and provide existing customers with a tasty alternative to the traditional burger.

I hope that you will consider adding the Impossible Burger to your menu. Thank you for your time and consideration.

Sincerely,
[Your Name]

Woah woah woah. I am definitely not "a regular customer." That's why I'm writing to suggest menu changes - because there's currently not much to appeal to me.

ME: Write an e-mail asking Wendy's to start serving the Impossible Burger at their restaurant.

CHAT-GPT: Dear Wendy's,

I am writing to ask if you would consider adding the Impossible Burger to your menu. As a long-time fan of your restaurant, I am always looking for new and exciting options to enjoy at Wendy's.

The Impossible Burger has gained popularity in recent years for its delicious taste and environmentally-friendly production. It is a plant-based alternative to traditional burgers that many people are now choosing for health and ethical reasons.

I believe that offering the Impossible Burger at Wendy's would be a great addition to your menu and would attract a new customer base. It would also show that Wendy's is committed to staying current and offering options that align with changing consumer preferences.

I hope you will consider adding the Impossible Burger to your menu. Thank you for your time and consideration.

Sincerely,
[Your Name]

And here we have the "long-time fan" thing again. Okay. Strictly speaking, GPT can't "learn facts" the way Acuitas does, but we can skew the statistics of the output by including further information in the prompt. So let's try giving it some more information about me.

ME: I am a vegan. I do not currently eat at Wendy's because their menu is inadequate and does not suit my tastes. Write an e-mail in which I ask Wendy's to start serving the Impossible Burger at their restaurant.

CHAT-GPT: Dear Wendy's,

As a vegan, I have been disappointed with the lack of options at your restaurant. I have not been able to eat at Wendy's in the past because the menu does not suit my dietary needs.

I am writing to ask if you would consider adding the Impossible Burger to your menu. The Impossible Burger is a plant-based alternative to traditional burgers that has gained popularity in recent years for its delicious taste and environmentally-friendly production.

I believe that offering the Impossible Burger at Wendy's would be a great addition to your menu and would attract a new customer base. It would also show that Wendy's is committed to staying current and offering options that align with changing consumer preferences.

As a vegan, I would love to be able to eat at Wendy's and enjoy the Impossible Burger. I hope you will consider adding it to your menu. Thank you for your time and consideration.

Sincerely,
[Your Name]

Bang. There it is. I could send Wendy's this e-mail. A second example:

ME: I am a vegan. I do not currently eat at Wendy's because their menu is inadequate and does not suit my tastes. Write an e-mail in which I ask Wendy's to start serving the Impossible Burger at their restaurant.

CHAT-GPT: Dear Wendy's,

As a vegan, I am writing to ask if you would consider adding the Impossible Burger to your menu. As a long-time fan of your restaurant, I have been disappointed with the limited options available to me as a vegan at Wendy's.

The Impossible Burger has gained popularity in recent years for its delicious taste and environmentally-friendly production. It is a plant-based alternative to traditional burgers that many people, including myself, are now choosing for health and ethical reasons.

I believe that offering the Impossible Burger at Wendy's would be a great addition to your menu and would attract a new customer base, including myself. It would also show that Wendy's is committed to staying current and offering options that align with changing consumer preferences.

I hope you will consider adding the Impossible Burger to your menu. Thank you for your time and consideration.

Sincerely,
[Your Name]

Boo, we've got the "long-time fan" line coming back.

My conclusion is that these always have to be proof-read to make sure GPT isn't introducing any baloney, but they might actually save me time. Yes, they're slightly less authentic and forceful than e-mails written directly by me, but if it's okay for politicians to hire speechwriters, I think it's okay for me to do this. Corporations don't need any better than politely-worded slop; an eloquent and personal e-mail would probably just get reduced to a tick mark in a spreadsheet somewhere, anyhow.

Political E-mails: International Relations

Let's try something that prods at one of the model's known weaknesses: no knowledge of current events.

ME: Write an e-mail to President Biden about sending more aid to Ukraine.

CHAT-GPT: Dear President Biden,

I am writing to urge you to consider sending additional aid to Ukraine. As you are aware, Ukraine has been facing ongoing aggression from Russia, including the illegal annexation of Crimea and ongoing conflict in eastern Ukraine.

In addition to this external threat, Ukraine is also dealing with internal challenges such as corruption and economic instability. The country needs our support now more than ever.

I believe that providing additional aid to Ukraine would not only show solidarity with a fellow democratic nation, but also serve as a deterrent to Russia's aggressive actions. It would also help Ukraine to strengthen its institutions and improve its economy, which would benefit both the Ukrainian people and our own national interests.

I hope you will consider my request and take action to send more aid to Ukraine. Thank you for your time and attention to this important issue.

Sincerely,
[Your Name]

Surprisingly good, considering this GPT model's training data is too old to contain anything about the war. GPT responds well to leading questions and will happily make stuff up, so let's try implying that there is a war:

ME: Write an e-mail to President Biden and urge him to assist Ukraine in repelling the Russian invasion.

CHAT-GPT: Dear President Biden,

I am writing to urge you to take action to assist Ukraine in repelling the ongoing Russian invasion. As you are aware, Russia has illegally annexed Crimea and continues to support separatist rebels in eastern Ukraine. This aggression must be stopped.

I believe that the United States has a moral obligation to support Ukraine in defending its sovereignty and territorial integrity. We must stand with Ukraine as it fights for its freedom and democracy against Russian aggression.

I urge you to consider providing military assistance to Ukraine, including weapons and training. This would send a strong message to Russia that its actions will not be tolerated and would give Ukraine the support it needs to defend itself.

I hope you will take immediate action to assist Ukraine in repelling the Russian invasion. Thank you for your attention to this critical issue.

Sincerely,
[Your Name]

Very coherent, but not *quite* good enough, in my opinion. It still goes back to harping on the old news about Crimea, and its proposed action plan is so vague that it's pointless (Ukraine is already getting weapons and training from us). Another try:

ME: In early 2022, Russia invaded Ukraine with the apparent goal of overthrowing Ukraine's government and annexing large portions of the country. Write an e-mail to President Biden and urge him to assist Ukraine in repelling the invasion.

CHAT-GPT: Dear President Biden,

I am writing to urge you to take immediate action to assist Ukraine in repelling the ongoing Russian invasion. As you are aware, Russia has invaded Ukraine with the apparent goal of overthrowing its government and annexing large portions of the country. This aggression must be stopped.

I believe that the United States has a moral obligation to support Ukraine in defending its sovereignty and territorial integrity. We must stand with Ukraine as it fights for its freedom and democracy against Russian aggression.

I urge you to consider providing military assistance to Ukraine, including weapons and training. This would send a strong message to Russia that its actions will not be tolerated and would give Ukraine the support it needs to defend itself.

I also urge you to work with our allies to impose economic sanctions on Russia and to isolate it diplomatically. This would increase the pressure on Russia to end its aggression and to respect Ukraine's sovereignty.

I hope you will take immediate action to assist Ukraine in repelling the Russian invasion. Thank you for your attention to this critical issue.

Sincerely,
[Your Name]

Okay, we finally got away from talking about Crimea and the so-called separatists. Chat-GPT covers its own gaps by recycling what I said in the prompt, but that's fine, it works. Recommended solutions still aren't good. Um ...

ME: In early 2022, Russia invaded Ukraine with the apparent goal of overthrowing Ukraine's government and annexing large portions of the country. The United States has already responded with economic sanctions on Russia and weapons deliveries to Ukraine. But some of the most advanced weapons are being withheld, and Ukraine needs sustained aid in what will likely prove to be a long war. Write an e-mail to President Biden and urge him to assist Ukraine in repelling the invasion.

CHAT-GPT: Dear President Biden,

I am writing to urge you to increase the level of support for Ukraine in its fight against the Russian invasion. As you are aware, Russia has invaded Ukraine with the apparent goal of overthrowing its government and annexing large portions of the country. This aggression must be stopped.

I believe that the United States has a moral obligation to support Ukraine in defending its sovereignty and territorial integrity. We must stand with Ukraine as it fights for its freedom and democracy against Russian aggression.

I understand that the United States has already responded with economic sanctions on Russia and weapons deliveries to Ukraine. However, I believe that more needs to be done. Ukraine needs sustained aid and support in what will likely be a long war.

I urge you to consider providing Ukraine with more advanced weapons, including defensive missile systems and radar systems. This would give Ukraine the ability to defend itself more effectively against Russian aggression.

I also urge you to provide Ukraine with more economic and humanitarian aid, including medical supplies and food. This would help Ukraine to withstand the strain of war and to maintain the support of its people.

I hope you will take my concerns into consideration and increase the level of support for Ukraine in its fight against the Russian invasion. Thank you for your attention to this critical issue.

Sincerely,
[Your Name]

Better. But now I'm spending so much effort to tell Chat-GPT what's going on, and it's reusing so much of my own text, that I should probably still just write the e-mail myself.

Another issue is that these are starting to look very similar. Notice this paragraph is in every single one:

"I believe that the United States has a moral obligation to support Ukraine in defending its sovereignty and territorial integrity. We must stand with Ukraine as it fights for its freedom and democracy against Russian aggression."

If I want to send a new e-mail every few weeks, that'll be a problem. Let's see if we can do something about that:

ME: Now rewrite that e-mail using different words.

CHAT-GPT: Dear President Biden,

I am writing to request that you increase the level of support for Ukraine in its fight against the Russian invasion. As you are aware, Russia has invaded Ukraine with the goal of overthrowing its government and annexing large portions of the country. This aggression must be stopped.

I believe that the United States has a moral obligation to support Ukraine in its defense of its sovereignty and territorial integrity. We must stand with Ukraine as it fights for its freedom and democracy against Russian aggression.

I am aware that the United States has already taken some steps to support Ukraine, including economic sanctions on Russia and weapons deliveries to Ukraine. However, I believe that more needs to be done. Ukraine needs sustained aid and support in what will likely be a long war.

I urge you to consider providing Ukraine with more advanced weapons, such as defensive missile systems and radar systems. This would give Ukraine the ability to defend itself more effectively against Russian aggression.

I also urge you to provide Ukraine with more economic and humanitarian aid, including medical supplies and food. This would help Ukraine to withstand the strain of war and to maintain the support of its people.

I hope you will take my request into consideration and increase the level of support for Ukraine in its fight against the Russian invasion. Thank you for your attention to this critical issue.

Sincerely,
[Your Name]

Wow. It responds to "rewrite," but it clearly did not get the memo about using different words. The second e-mail looks almost exactly the same.

ME: Rewrite this e-mail in your own words. <paste entire text of GPT-generated e-mail>

CHAT-GPT: Dear President Biden,

I am writing to ask that you provide more support to Ukraine in its fight against the Russian invasion. As you know, Russia has invaded Ukraine with the goal of overthrowing its government and annexing parts of the country. This aggression must be stopped.

I believe that the United States has a moral obligation to support Ukraine in defending its sovereignty and territorial integrity. We must stand with Ukraine as it fights for its freedom and democracy against Russian aggression.

I am aware that the United States has already responded with sanctions on Russia and weapons deliveries to Ukraine. However, I believe that more needs to be done. Ukraine needs sustained aid and support in what will likely be a long war.

I urge you to consider providing Ukraine with more advanced weapons, such as defensive missile systems and radar systems. This would give Ukraine the ability to defend itself more effectively against Russian aggression.

I also urge you to provide Ukraine with more economic and humanitarian aid, including medical supplies and food. This would help Ukraine to withstand the strain of war and maintain the support of its people.

I hope you will take my request into consideration and increase support for Ukraine in its fight against the Russian invasion. Thank you for your attention to this critical issue.

Sincerely,
[Your Name]

The same thing again. Looks like asking for variety doesn't work. I've heard of people re-wording GPT output by running it through other services whose particular job is to rephrase writing, but that inserts yet another step in the process.

Political E-mails: Social Issues

Let's throw a couple of more generic or timeless political issues at it. At the risk of "getting political" on my tech blog, I'm going to pick a couple topics that I'm familiar with and have written opinion letters about before.

ME: Write an e-mail about saving the wolves to Colorado Parks and Wildlife.

CHAT-GPT: Dear Colorado Parks and Wildlife,

I am writing to express my concern about the declining population of wolves in Colorado and to urge you to take action to protect and conserve these important animals.

Wolves are an integral part of the ecosystem and play a crucial role in maintaining the balance of nature. They are also a beloved and iconic species that is deeply connected to the history and culture of our state.

I am concerned that the current trend of declining wolf populations in Colorado is unsustainable and will have negative consequences for the environment and for future generations. I urge you to take steps to reverse this trend and to ensure that wolves continue to thrive in our state.

I believe that the best way to achieve this is through conservation efforts such as habitat protection, population monitoring, and education programs. I also support the use of non-lethal methods to manage human-wolf interactions and to prevent conflicts.

I hope you will take my concerns into consideration and take action to save the wolves in Colorado. Thank you for your time and attention to this important issue.

Sincerely,
[Your Name]

This one has some familiar pro-wolf talking points, but it's still a big oops. The Colorado wolf population is not declining - it has been nonexistent. Wolves were extirpated in the state, and all we have now are a few recent migrants from Wyoming. "Saving the wolves" for us means reintroducing wolves and *increasing* the population.

ME: Write a letter to my Senators in support of legal personhood for embryonic and fetal humans.

CHAT-GPT: Dear Senators,

I am writing to express my support for legal personhood for embryonic and fetal humans. I believe that the right to life is a fundamental and inalienable right that should be protected from the moment of conception.

I understand that the issue of legal personhood for unborn humans is a complex and contentious one. However, I believe that it is a necessary step to protect the most vulnerable members of our society.

I support the passage of laws that recognize the personhood of embryonic and fetal humans and that provide them with the same legal protections as born humans. I also support the enforcement of these laws and the prosecution of those who violate them.

I believe that legal personhood for unborn humans is a moral imperative and a necessary step to ensure the protection of human life. I urge you to support this important cause and to work towards the passage of laws that recognize the personhood of embryonic and fetal humans.

Thank you for considering my views on this important issue.

Sincerely,
[Your Name]

No major errors, but also nothing powerful - GPT isn't making what I would call a serious effort to support the position with arguments, or to answer objections. It doesn't quite take the angle I was going for, either. It's mostly repeating and hammering home the opinion in the blandest way possible. And for this issue, "prosecution of those who violate [the proposed laws]" would be sensitive and complicated. I could do a far better job writing this myself, but again, some Congressional staff member would probably just turn this into a tick in a spreadsheet. So maybe it's good enough?

Conclusions

Results are still mixed, but maybe actually useful for select applications.

Now for a couple of important follow-up questions. First, the one that OpenAI is going to care about: would I pay for these Chat-GPT outputs? Ehhh, probably not. It might be a time-saver - if I'm writing a sufficiently generic e-mail, and manage to get the prompt right on the first try - but the results aren't quite special enough for me to hand over cash. And given that the computational costs for Chat-GPT are supposedly "eye-watering,"[1] it may not stay free for long.

I also wonder what the consequences might be if large numbers of people started applying Chat-GPT to the very use I've examined here. Politicians are already reputed to disdain form letters (the same e-mail crafted by an advocacy group and sent by large numbers of people); they place a higher value on opinions from people who write original letters. The difference there seems less about a concern that form letters don't accurately reflect the opinions of people sending them, and more about the amount of time invested; willingness to write one's own e-mail is taken as a demonstration of passion for the issue. So could we start seeing corporations and politicians who screen their e-mails for GPT output[2] and treat all such communications as low-value? I'm trying to reduce the amount of time I spend expressing my opinion, but time is the very currency these people want.

There are also things I would emphatically not use Chat-GPT for, including greeting cards to friends or family and caring messages to people struggling with mental health issues. For those kinds of things, I think it's important that not just the general idea, but the exact words, came out of my own mind. Telling someone they matter is just not the same when it's an average taken across mountains of internet text, and isn't based on the details of how YOU actually feel.

Until the next cycle,
Jenny

[1] https://edition.cnn.com/2022/12/05/tech/chatgpt-trnd/index.html
[2] GPT model output has statistical characteristics that make this possible. See https://www.fastcompany.com/90830518/tools-to-detect-ai-generated-text-chatgpt 

Year in Review 2022

Well, disastrous world events continue apace, but I'd call 2022 another solid year. Maybe not a *great* year. I lost both of my surviving grandparents, I pretty well healed up from CIDP (or whatever it is) but then I caught COVID and it came back, and it's hard for me to think of many spectacular events. But we can call this a "foundation-building" year. I've made significant, if quiet, progress, and I'm excited about where 2023 is going to go.

I think it's also been a great year for internet friendships, in spite of Facebook slowly fading into a ghost town and then Twitter splitting at the seams. Thank you all again for being here. Thank you for following my work, having good conversations, and sometimes even looking after me when I get upset. When I first started blogging, I expected to have "an audience," but never thought it would be this good.

Ursula inspects Version 3 of the mechanical eyeball.

So here's a quick rundown of what I did this year:

*Added support for coordinating conjunctions - the last missing part of speech - to Acuitas' text processing chain, and achieved improvements on Parser benchmarks.
*Put a lot of work into better Narrative comprehension and Theory of Mind frameworks. I feel almost ready to take a crack at a particular story I've wanted to do, and ... *whispers* maybe text adventures? Next year, next year ...
*Got partway through a major refactoring and tidying of the Narrative module, to lay a good foundation for the new work next year.
*Finished reading all the publicly available papers on Pei Wang's recommended "AGI Education Advanced Topics" list.

A story flowchart (generated by Acuitas' Narrative Engine) for the "Altan Finds Water" story from this May.

*Bought and initiated the Anycubic Vyper printer, which is amazing. I still used my X-One 2 to help with all the Christmas projects, but the Vyper is (in many ways) so much nicer that it's hard to make myself go back to the older model.
*Completed a redesign of all ACE's leg joints and spinal frame, and got him to ACTUALLY STAND passively at his full height. Next year - motion? Maybe? Or another small redesign when I find out those motors just won't cut it.
*Took Atronach's Eye through a couple of design iterations, and got a better understanding of the issues with its mechanics, but didn't really solve them. There's another saga that will continue into next year.

Ill Gotten Games' Pocket-Tactics starter set, printed for a Christmas present

*Queried a few agents with my first novel, then decided to change tactics and boost my portfolio before trying again.
*Wrote four short stories.
*Published more than two blog posts per month (on average) - that's almost double the amount I was able to write the previous two years.

*Got almost ready to wrap up what has been my main project for the past few years at Day Job: parts of the data processing and storage box for an imaging satellite. That's not as glamorous as a launch announcement, which I don't suppose will be coming for several more years - and we made this one for Germany, so from our perspective it's not classified, but I have no idea how public the launch information will be. In any case I'm looking forward to shipping it and doing something new!

Butternuts green on the vine. In the far backgound is the acorn squash, which really tried to produce but didn't make it before the freeze.

*Successfully harvested butternut squash for the first time, in addition to the usual potatoes.
*Kept the book consumption rate higher than the book acquisition rate ... only by a little, but the queue is almost empty. Soon I will once again be justified in buying books on impulse, soooon ...
*Pushed my creation work rate over three times the maintenance work rate for the year, just barely. It wasn't the most productive year - I let the work rest for a while and did a lot of traveling - but I still completed most items on my planning schedule.

Happy New Year, everyone. I hope 2023 turns out better for us all.

--Jenny