# 18. Representations: Classes, Trajectories, Transitions

**Channel:** MIT OpenCourseWare
**Source:** https://www.youtube.com/watch?v=bQI0OmJPby4
**Transcript page:** https://www.withtranscript.ai/video/bQI0OmJPby4

## Chapters

- 0:14 — Chapter 1 (0:14)
- 7:41 — Chapter 2 (7:41)
- 9:52 — Chapter 3 (9:52)
- 12:12 — Chapter 4 (12:12)
- 13:47 — Chapter 5 (13:47)
- 17:22 — Chapter 6 (17:22)
- 18:48 — Chapter 7 (18:48)
- 20:01 — Chapter 8 (20:01)
- 24:02 — Chapter 9 (24:02)
- 25:33 — Chapter 10 (25:33)
- 26:04 — Chapter 11 (26:04)
- 28:45 — Chapter 12 (28:45)
- 31:07 — Chapter 13 (31:07)
- 32:09 — Chapter 14 (32:09)
- 34:52 — Chapter 15 (34:52)
- 35:21 — Chapter 16 (35:21)
- 38:05 — Chapter 17 (38:05)
- 40:20 — Chapter 18 (40:20)
- 41:09 — Chapter 19 (41:09)
- 43:18 — Chapter 20 (43:18)
- 44:00 — Chapter 21 (44:00)
- 46:30 — Chapter 22 (46:30)
- 48:44 — Chapter 23 (48:44)

## Transcript

**[0:14]:** we might wonder a little bit about the nature of human intelligence and we might reflect a little bit on the kind of intelligence we've been talking about in the past few weeks it's been an intelligence of sorts those programs support vector machines boosting they can do really smart things but the peculiar thing about systems that use those methods is that those systems don't have any idea about what they're doing they don't know anything so they don't give us very much of an insight into the nature of human intelligence and after all we'd like to have a model of human intelligence because let's face it with the smartest things around so there are lots of ways to approach that question and we'll approach that question first from a evolutionary point of view some scientists believe me for instance that we have a family tree that looks about like

**[1:25]:** this

**[1:29]:** too small to see much of that but the main point is that we haven't been around very long we humans have been around maybe 200,000 years and the dinosaurs died out 60 million years ago so in the blink of an eye we seem to have more or less taken over when you look at that family tree and a scale where you can see it one of the characteristics is increasing brain size there we are in the left chimpanzee on the right plainly mostly mouth not too much brain in there that one down below is a reconstruction of one of those pivots type my petal Apes from about four million years ago or so so we became bipedal a long time before we had much of a brain so we might think that well maybe brain size has got a lot to do with it and I suppose it does so

**[2:34]:** we can plot brain volume of the our ancestors versus time so the picture I just showed you was from about three million years ago I guess and then up on the upper right hand corner oh that's not just us that's also the Neanderthals their brains might have been slightly bigger than ours so it isn't just brain size here's what that guy looks like that's a Neanderthal of course on the left and that's one of us on the right some conspicuous differences big heads the ribcage is kind of conical in shape yet a large pelvis people out make liked to make a lot of speculations about how they must have moved around but one thing is plain they didn't amount to much they could make stone tools but their stone tools didn't change much over tens of thousands of years and that was pretty much the story with us too until something happened probably in southern

**[3:38]:** Africa probably in a group of individuals maybe less than a thousand what's the evidence for that the evidence for that is mostly comes from DNA studies with a lot of probabilistic assumptions and Montecarlo simulations but it seems that among the competing hypotheses for how we came to populate the world it seems that there was a group of us Homo sapiens in southern Africa that got something that nobody else had and and and the highest probability scenario is that we quickly took over that population of Homo sapiens dominated the rest went out of Africa and within the blink of an eye did that sort of stuff what's that sort of stuff those are on those two paintings are from Moscow about 25,000 years ago paleoanthropologists like catters all take that as plain evidence that there was symbolic shot and the people who are around at that time us almost a pians the head is a carving off of mastodon tusk of a woman 25,000

**[4:49]:** years ago also plainly symbolic people were making a lot of jewelry and doing self adornment the Neanderthals never seem to do that that jewelry making seems to have gone back to southern Africa maybe 70,000 years ago people were puncturing seashells and using them as necklaces apparently so something happened and the paleoanthropologists who write fascinating stuff don't quite know how to talk about other than say that looks like we became somehow symbolic and somehow that has something to do with language so if you talk to me Tom ski he will say let me get this precise this is nearest quotation I expect as I can get he thinks it was the ability to take two concepts and put them together to form a third concept without disturbing the original concepts and without limit and each part of that's important see without limit part is what separates for us from species that might be able to do

**[5:50]:** that a little bit but we can do it without any apparent limit so that's a linguist speaking and he talks a lot about the merge operation and combinators in languages using terms that are foreign to us they don't use the term Combinator it's a kind of computer science term but whatever it is it seemed to happen about that time it was didn't happen slowly in proportion to brain size it seemed to happen all of a sudden in consequence of a brain that had grown big enough to be an enablement but the capability was not pulled up not what pulled evolution in that direction so I believe that whatever that was that capability enabled humans as humans to tell and understand store and that's what separates from us from the other primates that ability to symbolic ability what ever it is enabled storytelling and understanding and that's what all of education is about and that's why

**[6:54]:** our species is special so we're going to talk about today is some something you might think of as an instantiation of that hypothesis one way of thinking about it and it's a way of thinking about what the linguists would call the inner language it's not the language with which we communicate it's the language with which with which we think which is closely related to the language which with which we communicate but may not be quite the same thing so many of you are bilingual Christians bilingual if you ever had the experience of remembering that someone said something to you but not remembering what language they used

**[7:41]:** how about you Sonia if you had that experience well so yeah experience of having landed if you ever had that experience of having remarried remembering some conversation but not remembering the language in which it was cast

**[8:01]:** usually no that's a common that's a common view you remember something was said that there was a conversation that had some content but if it's with a speaker of your own language and your embedded in another place you often don't remember what language the conversation was in is that right Juana remembers things like that not sure so sometimes sometimes you don't have that confusion she says because you always speak to particular people in a particular language but many people report that they have that experience of not remembering which language something was set in well okay so what are we going to do we need an inner language and maybe we can start just by saying let's have let's have something that looks sort of familiar to a sweet we have an object and it's supported by some other objects so those are support relations

**[9:01]:** that's one example of what we might call a semantic net it's a network it's got nodes and links it's got it's got it has a meaning that's where the word semantics comes from or we might have another example it looks like this there's Macbeth there's Duncan and Macbeth murders Duncan and we also know somehow that there's a kill involved as a consequence and that ultimately Duncan has a property

**[9:52]:** and that property is the property of being dead so there's another semantic net recording something that happens in Shakespeare's Macbeth plot now we can decorate that a little bit so as to get a couple of other concepts and play first of all the thing we've got already is we've got combinators

**[10:21]:** then come in a Taurus the fancy name for those links that connect the notes another thing we've got is an opportunity for connecting the links themselves so the murder sort of implies the kill and the kill leads us to conclude that the victim is dead so that is treating the links themselves as objects that can be the subject or object of other links so we call that process reification

**[11:04]:** now in artificial intelligence cementing Nets we're all over the earlier work but you know if you have a big network that covers the wall you need some way of sort of putting a spotlight on some pieces of it so Marvin Minsky put a lot of technical content into that idea and created the notion of a notion that deserves another call a color here he suggested that we need a localization process so we have frames so called frames or templates and a frame for this murder action might be that there is a murder frame that has an agent and has a victim and the agent is Macbeth and the victim is dumpin so that's a way of putting a kind of localization layer

**[12:12]:** on top of what we've got so far later on I'll add sequence to that list so this is where it rested for a long time and in some sense still rests there because this as soon as you've got as soon as you got combinators and reification you've got something that's pretty here that's pretty much universal you could do anything with it the trouble is it's sort of down at the bit level it's like assembly code it doesn't have as a concept enough organization to help you go to the to the next level of achievement there's also a little problem here that deserves also some mention and that is that we have over this whole thing the problem of parasitic semantics

**[13:05]:** CITIC semantics a

**[13:14]:** kind of ugliness that surrounds this whole concept because when we look at a diagram like that and we say Oh Macbeth murdered Duncan that means Duncan's the victim we know that there must have been a motive maybe Macbeth one of the became we all we know all that stuff and there's a tendency to project that knowing into the machine if you've got to play with your telephone please leave so if

**[13:47]:** if we project meaning into that that's that's that's our understanding that's not the machines understanding so much of the meaning can be said to be parasitic it's a it's a parasite where the parasite and we're projecting the meaning into that thing putting that diagram into some machine form doesn't mean the machine knows anything it might be able to conclude some things but it's understanding is not grounded in any kind of contact with the physical world so we have to worry a lot about that and we were we were philosophers would stop there and go off and write a few books on the subject but we're not philosophers so we're going to just mention the problem and go barreling ahead so we need to use this notion of semantic net and we have to ask ourselves some questions about what elements of the inner language are most useful in the end it might be very complicated

**[14:36]:** but here's usefulness number one the notion of classification so we know about stuff and we know about for example pianos

**[14:51]:** and we know about tools and we know about max but we know about those things on different levels so when I say I'm thinking about a tool do you have a very good image of what I'm talking about the answer has to be no because the notion of the tools very vague so it's hard for you to form a picture of what that's all about another hand if I say I'm thinking about Mack well this is interesting because there's lexical ambiguity there you don't know if I'm talking about the the Apple type Mac or the Apple type Mac or should I say the fruit or the computer so there's lexical ambiguity there are two levels or more but let's fill this in a little but if I know about I know I'm talking about a piano you can form a picture of that so that seems to be in a more detailed level where you

**[15:51]:** can do hallucination at a higher level you have just a musical instrument

**[15:57]:** and I can give you a tool to think about by writing hammer and if I'm going to have a Mac it's going to be an apple and in this case I want you to think about a fruit and down here I can be more specific about these things too I can add a slight refinement of detail and say I'm thinking about I'm thinking about one of these do you know what this is yeah so it's not a mere hammer it's a ball-peen hammer and some circles is called a lady's hammer I don't know why but it's what's this for most people buy it mostly because it's small and lightweight but in fact it's for metalworking just we're taking a piece of sheet metal and pounding it out into an ashtray or something or four-cd rivets it's a metal workers hammer so you might not have known about that before but now at least you have a word to hang that knowledge on it's a ball peen hammer

**[17:05]:** so we have various levels here going from very specific to very general and we could even go to a level of specificity for pianos by saying that we've got two bows indoor fur

**[17:22]:** why is it boughs and offer special and it is like a Baldwin or something may not yoga

**[17:36]:** lots of piano types what's special about place you don't know because you unless you play the piano and probably unless you're a serious piano player you don't know that a Bose endure for Ariel you know

**[17:52]:** it's got some extra keys at the bottom and most people don't know that unless they're serious about the piano some professional piano players when they're confronted with the bows and or fur have to cut have to have someone cover those keys because it screws up their peripheral vision and I hit the wrong key because I'm not used to having those extra keys at the bottom so that's that's a little detail about the bows on door first so you can make kind of graph and you can say let's go from who very general

**[18:27]:** to a sort of basic level to a specific level and so it is this it is the case in human knowledge that that graph has a tendency to look sort of like this

**[18:48]:** so here's tool here's hammer here's ball-peen so that level where you have a big jump that's the general - basic level transition so that basic level is probably there because that's the level at which we on which we hang a huge amount of our knowledge we know a lot about pianos and it all seems to be hanging on that word piano which gives us power over the concept so that's example number one of a of an element of our inner language the ability to assemble things into hierarchies like that and to hang knowledge about those objects into the in that higher on the elements in that hierarchy

**[19:34]:** but given that you have elements in the hierarchy to talk about how do you talk about them well I like to consider the possibility just for the sake of illustration that you're thinking about a car crashing into a wall so you've got the things to think about like the speed of the car

**[19:54]:** the distance to the wall

**[20:01]:** and the condition of the car and you've got the period before the crash during the crash and after the crash so you might want to think about how to talk about those objects in those three time periods so

**[20:24]:** we can do that with a vocabulary of change and we do that because we believe that most of human thinking is thinking about change causing change and that flies in the face of what we learn as engineers because in engineers in engineering we learned about State and once you know the state of a system you know everything you need to know in order to predict the future the trouble is in our heads thinking about everything there is in the world including the current phase of the Moon is too much stuff so mostly our thinking we think is is hinging on the idea that change leads to change so that's why we have a vocabulary of change so that so in the period before the crash the speed of the car is not changing there's a little notation for not change no Delta the distance to the wall that's decreasing the condition of the car that's

**[21:25]:** not changing but then the car hits the wall so the speed of the car disappears the distance of the wall disappears and the condition of the car will change dramatically finally after the crash is over the speed of the car does not appear the distance to the wall does not change and the condition of the car also does not change so that's hinting at a vocabulary of change and its use which will be the second element in our development of a vocabulary of ways in which we might have constructed our in our language so this is a particular idea that's classification this is transition and in any system that purports to understand stories there'll be a heavy emphasis on this notion of transition and we believe that it's a I think that that vocabulary has to have decrease increase change appear and disappear so there are ten things you can have in such diagrams I've

**[22:40]:** done five that's because for each of those there's a not variation on that so vocabulary ten things can go a long way toward helping you to describe things that are in process of changing or making transition and we have a lot of those words in our vocabulary we use those words a lot in our vocabulary they seem heavily connected with vision our friend appeared the cat disappeared the speed increased so this is a description of a crash

**[23:15]:** in terms of those kinds of elements now I say to you how does a camera work well I could say a camera works because a photon hits a photo verse fo toda a photon crashes into a photoreceptor so when I say a photon crashes into a photoreceptor why am i saying that and how does it help I'm saying that and it helps because it's the same pattern of change you already know about wanting to talk about car crashing into a wall how does that work the speed of the photon

**[23:55]:** the distance of the photon

**[24:02]:** to the receptor and the condition of the receptor

**[24:12]:** so analogies like that are the very much at the core of what we think about all the time in any event there is there is representation number two number one is class number two is transition and now you're ready for number three which is trajectory linguists who study sentences often talk in terms of fundamental patterns that seem to be in a lot of what we say and a lot of what we say is about objects moving along trajectories so we can talk about a trajectory frame

**[24:59]:** and a trajectory frame will have elements like this it has an object moving on a trajectory that ends up at a destination

**[25:10]:** it might start out at a at a source it's probably been arranged by some kind of agent and their agent may assist himself in making the motion happen with some kind of instruments

**[25:33]:** there might be somebody helping out over here a co agent

**[25:40]:** well what else could we have beneficiary someone who is helped out by the action

**[25:54]:** sometimes the motion is arranged by a conveyance

**[26:04]:** so these are a lot of slots frame like slots in descriptions of actions many of which involve motion along a trajectory we have a tendency in language to decorate these things in one way or another depending on the language in some languages the decoration is by way of position in a sentence in English it's often by way of a preposition that's used to help zero in on the role of a particular role of an object in the trajectory scenario so if I say I baked a cake with a friend there's a with preposition if I bake the cake for a friend the friend is the beneficiary if I baked a cake with an oven that's an instrument the object may be moving to a destination from a source and if I'm going to New York by train I put a bi on top of that if the agent is

**[27:14]:** an in subject position I would say something like oh the all the work was done by a student so those prepositions have a tendency to help us to zero in on the actual role of particular objects in this whole package this whole frame so this is number three and there's a variation on this in which there's no actual trajectory in which case we'll just call that a roll frame because if there's no trajectory you can still have things such as an instrument a co agent and a beneficiary so now we've got three representations you might say well what good are they and you can determine what good they are these days because it's easier to go over established corpuses and say what fraction of those of the sentences in such a corpus involve a classification or transition or a trajectory the most well-known of these is the so called Wall Street Journal corpus it has

**[28:10]:** fifty thousand sentences in it drawn from some period of time all the language syntactical language types work with with that corpus a great deal and we worked with a little bit too to see what fraction of the of the sentences or what the density of trajectories and transitions are in those sentences so I have to say that a little more a little carefully because the finding is that in 100 sentences

**[28:45]:** you'll find about 25 transitions or trajectories so they're very densely represented they're often very abstract prices rose the economy went to someplace but there are they're still words the denote transition era trajectory but of course once you have all this stuff that you have a we didn't have a desire to put it together and so the next thing we need to talk about is story sequences so a story sequence can be a single sentence and I want to illustrate that with the one of my favorites

**[29:34]:** here's the sentence

**[29:47]:** there I think I've chosen gender-neutral name so it's not a gate trouble so that by the way don't call myself Pat because I decided when I was 18 years old that Pat is a unit of measure for butter in any case Pat comforted Mary do you have a image of what happened probably not a very firm image you know that Pat did something but you don't know exactly what nevertheless when Pat comforted Chris you can construct something that looks like a roll frame because the roll frame for that would have an agent and that would be Pat there's an action we're going to put a question mark in there because we don't have a very firm image of what the action is any of it we're building a roll frame like so the object is Chris oh let's you know that is the object

**[31:07]:** so anything else we can say yes we can probably say something more something else comes to mind when when you see Pat company press there was there's a sort of result and the result is a transition frame

**[31:27]:** and the transition frame involves an object which is Chris and Chris has a mood

**[31:44]:** which presumably is improved it goes up did you have something like

**[32:09]:** wandering into a very interesting area having to do with when you think about this in another way and think of it as something moving along a trajectory comfort is moving if not from Pat at least to Chris and that's a very important kind of observation because what what it suggests is that there can be utility and thinking of things in multiple ways multiple representations Marvin Minsky has a wonderful ristic phrases which is if you can only think about something in one way there's you have no recourse when you get stuck so multiple representations mean you have multiple ways of gathering regularity from the world and collecting it and therefore that will make you smarter so yes you could do it that you could do that and that would be a complement to what I'm doing now well let me continue what I'm doing now so what have I done I've got a roll frame and

**[33:03]:** a transition frame and the transition frame is the target of the result slot in the in the roll frame now we can modify this a little bit and maybe we want to say instead of comforted terrorized

**[33:29]:** and how would that change things we don't know exactly what Pat did so the action remains unknown the agent and the object are the same but the result here is presumably that the moved went down with just what we've got so far we can answer a lot of questions by the way once we've got the the sentence understood in these terms we can say who did the thing who which is all about and the answer is Pat what did Pat do comfort terrorize who did he do it to was the object it's Chris what was the result Chris felt better Chris felt were felt worse so these were these representations already give us a questioning answering capability that makes the it makes that makes for an understanding of the sentence still we haven't been very specific so our next step takes this the same sentence in a more specific direction so here's the way that

**[34:32]:** goes kissed Chris now you begin to get a sense of what's going on you can form a mental image so that you can have a hallucination

**[34:52]:** and a hallucination will also be a kind of frame but in this case it'll be a trajectory frame

**[35:03]:** and the object would be I don't know Pat slips

**[35:15]:** and the destination

**[35:21]:** would be Chris's lips I don't know is that right it depends it depends have you all form a picture of what's going on here so it will be different depending on whether Chris is Pat's girlfriend or if Chris is Pat's daughter or if Chris is a frog and Pat is a prince yes I guess the way the story usually does so somehow we have in our heads all kinds of libraries that help us to form mental pictures of things when we see things like hist so one final one just to show the variety we could say that Pat stabbed Chris

**[36:19]:** what changes let's see in the case of kissed the mood is going up in the case of stabbed the mood is going down you can also probably say that the health is going down and the destination is Chris's body and the object that's moving is Pat's knife

**[36:46]:** but you get the same pattern with both of those sentences so both of them involve a sequence that starts off with the action leads to a transition and a trajectory and those are all arranged in a line and that line is something that gives us a lot of power over the situation relative to a semantic net so I'm going to decorate that one more time and say that another element we get out of our internal language is sequence another element we need in order to have and anything that looks like an account of an inner language is sequence because if you think about things being arrayed in a vast spreading network it's hard to deal with them but if you think about things being arrayed along a line in a sequence of actions or events like so then that imposes enough constraint to get a handle on what's going on so what we're going to

**[37:47]:** call this is the representation and I guess I'm up to one two three four this is a representation of story sequence

**[38:05]:** so even though that's a kind of micro story it's still an example of a story sequence because we get the power out of it by arranging everything online now do you understand sighs if you play a music musical instrument of how dependent we are in sequence so if you play a musical instrument you probably know how difficult it is to start replaying a piece of music from the middle of a measure you have to go back to at least the beginning of the measure and probably to the beginning of the phrase if not the beginning of the piece so our memory seems to be at least in music very rooted in the idea of sequences and that's often true in storytelling too we have to go back to the beginning of a at least a scene because somehow these things are arranged in in sequences that form somehow useful usefulness out of there sequential Mis

**[39:02]:** so there's one more thing we can talk about and that is the idea of not just the idea of how these sequences are constructed and what they're constructed out of but we can also talk a little bit in terms of libraries of stories

**[39:25]:** and when we talk about libraries of stories we can think about kind of the sort of standard standard stories that we have and how they're arranged and how we can know a lot about something by what it's superclass's so it's a variation on the theme of learning stuff from the super classes so here's here's an event frame and then there's a addition to event frames that are disaster frames and then there are party frames and you know parties and disasters are both events and when we talk about disasters we know that they in turn break up into a variety of things we have earthquake disasters

**[40:20]:** and maybe we have hurricanes

**[40:25]:** and in the party world we have I know birthday parties

**[40:34]:** and we have weddings and each of these types of frames invites us to fill in particular slots so if we're reading a newspaper story about a wedding we know that we're going to be learning the same sorts of things we would learn about any party except there is the addition additional information that we expect that says something about the bride and the groom

**[41:09]:** if we have a raw event and we don't know anything more about it than that there's a time in place

**[41:19]:** if it's a disaster frame gets a little drizzly over here but if there's a disaster frame frame where you might have the fatalities and how much how much it costs if it's an earthquake frame we need to know the magnitude of the of the quake and the name of the fault if it's a hurricane we have the category and the name so each of these things up there can be viewed as as a No not just an example of something but as a new frame all by itself so as we as we mature we have these first four things as building blocks and then we we educate ourselves and we get all those kinds of frames that help us to understand the world but how to fill those in from a newspaper story can be sometimes quite a challenge actually the worst thing to understand is children's stories because and this was determined experimentally when people try to understand children's stories because it turns out the children's stories are not simpler than the stories we write

**[42:46]:** for adults in many cases they are they're harder if you reached about Shakespearean plots it's all about intrigue murder jealousy greed but when you're talking about a children's story it can be about anything and worse yet the children's story often raises problems that you don't see in newspaper stories let me illustrate that for you

**[43:18]:** read that story

**[43:24]:** you have no trouble figuring it out of course but think about app or machine it's struggling to understand anything what's going to be the problem it's going to have trouble figuring out those pronoun antecedents look at them they're complicated

**[43:50]:** you

**[44:00]:** they're the pronouns one of them wanted to buy a kite he has one he said he will make you take it back so that's pretty hard actually the principle here that I'm driving out is when you have a new story or any story if you have an old story but you want to read it where you want understand it quickly the new instantiation of it quickly you need to be sure that if you're the storyteller you don't add to the burden of the understanding and a syntactic difficulty so that's an example of telling a story with additional syntactic difficulty no newspaper journalists would ever write the story like that here's how they would write it they would even they would even give you a clue that there's certain information you're never going to get like who told the story that's that reliable sources business so this brings us to the to the to the final

**[45:00]:** bit that I want to deal with today and what I'm going to do is I'm you came into this class today as a 97 pound writing weakling you're about to emerge as a 250 pound mountain of a writer because I want to tell you a few tricks that will make you usually better as a writer especially especially if you're Russian or German and here's how it works rule one

**[45:44]:** because they place additional syntactic burden on the understanding of the story by the reader so here teaching so are you telling somebody about some difficult new technical idea the last thing you want them to do is to burden their syntactic processor with figuring out pronoun I students so don't use pronouns and your writing technical writing at least will be much clearer by the way why does this especially apply to Germans and Russians is it because is this is this a ethnic Marjan slur or it's just the fact about their language it's a fact about their language what is the fact why can they get away with pronoun usages that we cannot get away with in English yes

**[46:30]:** because if you have all of your all of your nouns decorated with with gender that reduces by three the the the potential for ambiguity of the pronoun antecedent so you'll frequently find that German Russian writers will have pronouns all over the place that are perfectly clear to them because of gender they're uninterpretable by English speakers when translated because we don't have the gender to help us zero in so these things are all have to do with minimizing extra superfluous gratuitous unnecessary burden on the reader number two is don't use former

**[47:12]:** or ladder you see that those words use frequently technical writing and guess what no human being ever encounters those words without having to stop and go back to figure out what they refer to so that's another example of not placing an unnecessary syntactic burden on the reader and finally don't call a shovel

**[47:39]:** a spade there's a there's a habit probably instilled by well-meaning but Miss advised - teachers that you shouldn't repeat words and so people will go to great lengths to use some different work the problem is that the reader doesn't know if the shift in the word is deliberate and attached to some subtle meaning shift or if it's just adhering to some high school teachers admonition against using the same word again so you don't want to say oh the right way to dig this particular hole is with the Spade and then switch to a shovel because the reader can't tell if it's deliberate accidental or a consequence of just the desire not to use the same word so this is how you can with some very simple mechanisms grounded in AI you can actually make yourself into a better writer by avoiding those kinds of things that put a unnecessary syntactic burden on the people are reading your

**[48:36]:** stuff

**[48:44]:** you
