Is AI About to Automate Every Office Job? (Not a Chance)

Added: 2026-05-06

[00:00:00]Back in February, Microsoft chief executive Mustafa Sulleman sat down for an interview with the Financial Times.

[00:00:08]During it, he made the following extraordinary claim. I think that we're going to have a human level performance on most if not all professional tasks.

[00:00:19]So, white collar work where you're sitting down at a computer >> either being, you know, a lawyer or an accountant or a project manager or a marketing person. most of those tasks will be fully automated by an AI within the next 12 to 18 months.

[00:00:35]>> Now, if this prediction is true, then we're just a year away from one of the most sudden and calamitous economic shifts in the history of modern economics. I mean, worldwide, the knowledge and technology intensive industries produced over produce over 10 trillion dollars of value per year and make up more than a third of economic activity here in the US. So if basically all of this could be replaced by compute and this was going to happen by next spring, it would make the industrial revolution seem glacial by comparison.

[00:01:05]It would be the economic equivalent of the asteroid that killed much of life on Earth, including the dinosaur. So is it possible that Sullean is right? And if he's not, what's a more realistic understanding of what AI will and will not be able to do in the workplace in the near future?

[00:01:24]Well, it's Thursday, which means it's time for another AI reality check episode. So, this is a great opportunity to dive deeper into Sulleman's claims.

[00:01:34]Now, I have a lot to say on this topic, including if you make it all the way to the end of this episode, a little conspiracy that I uncovered when I was doing research on this topic, so stay tuned for that. Um, but we have a lot to get to. So, let's get started. As always, I'm Cal Newport and this is Deep Questions, the show for people seeking depth in a distracted world.

[00:02:03]All right, as you may have guessed, I'm going to argue here today that Mustafa Sulleman's claim is not accurate. Now, I have three major reasons to propose why he is inaccurate in his claim. And I'm going to present these three reasons in order from least technical to most technical. Now, to be clear, I'm not trying to be like Dr. AI skeptic guy here, right? I mean, obviously people are finding uses for LLMs in the workplace, even if they are nowhere ready to replace all knowledge work jobs. So, I'm hoping that as I get to the uh end of these three reasons and before I get to that conspiracy I promised you at the end that I'll be able to review the ways that these tools actually are being useful. what are the actual parameters of where LLMs are or will continue to make a difference in knowledge work in the near future. So, we're going to give some positive information as well towards the end of this episode. All right, but let's dive in. I want to start with my first reason why Sulle Min is likely wrong in his claim.

[00:03:03]That reason is other tech leaders don't agree. So before we get into the details of what LLMs can or cannot do, I want to emphasize that Sulleman's idea that we are 12 to 16 months and given that he said this in February, that means basically a year away from all knowledge work jobs being fully automatable. So he said fully automatable by AI. That claim is an outlier compared to what other leaders of major AI companies have been saying. So for example uh before Sulleman grabbed the the crown of most doomer about AI job impacts last uh in February Dario Amade had been the most pessimistic among the CEOs on multiple occasions. The claim that he made is that AI will replace up to 50% of entry-level knowledge work jobs within 5 years. That's the claim that Amade has made several times now. Now look that that's not great news either but it's actually much less dramatic than what Sulleman was predicting. Think about it.

[00:04:06]It's a longer timeline. Amade is talking up to 5 years. It's only affecting entrylevel jobs, right? So he's not talking about all knowledge work jobs but just entry- level jobs. And he's not talking about most jobs. He's talking about 50%. So on all scales we might measure Amade's claim. It's much less drastic than what Sulleman was saying.

[00:04:27]If we turn our attention to Nvidia's Jensen Wong, uh he's much more aggressively against the type of claim that Sullean or even Ammedday is making right now. Um he doesn't really see AI fully automating many jobs at all. And in fact, at a recent event at Stanford, Wong argued that making these predictions about mult large swaths of the economy being automated is actually anti-productive, if not just straight up false. So I want to play a clip of Jensen Wong uh at a recent event making the opposite claim as men.

[00:05:00]>> First of all, I I think this the narratives of AI destroying jobs is not going to help America.

[00:05:07]>> Yeah.

[00:05:08]>> First of all, it's just it's false.

[00:05:10]>> Wong goes on in this uh this event to argue that what we're more more likely to see is AI tools being integrated into work like we did with computer tools in the 90s and early 2000s. It will change what jobs look like. it'll change the day-to-day, maybe what tools you're touching, but is not going to wholesale replace large swaths of the economy. He notes in that Stanford appearance that the engineering teams at Nvidia use a lot of AI tools and they're busier than ever before and they're hiring more engineers than ever before. So, he doesn't see AI tools as a job destroyer, but instead as a job changer. All right, so let's move on now to the second reason why Sulleman is likely wrong in his prediction.

[00:05:51]We are not seeing enough progress. All right. So, Sulleman says we're basically a year away from most knowledge work jobs being fully automated. If that was true, we would need to be seeing major rapid advances in LLM technology to keep us on such a uh ambitious trajectory.

[00:06:11]But it's simply not what is happening.

[00:06:14]Now, look, this might sound crazy at first because you're bombarded with news constantly about AI and AI is this and AI is that and we're scared of this and it gives you this general impression that aren't things just moving really fast in the AI world. But if you cut through the PR and the hype, here's what you will actually discover if you follow this technology closely. Since roughly late 2024, progress with newly released LLMs from the frontier uh AI companies have been making steady but not particularly fast progress.

[00:06:45]And instead of seeing the type of immediately obvious functional improvements that we got used to in that era when we went from GPT2 to GPT4, most of the improvements we're seeing from model to model are largely being captured in benchmarks. So charts of numbers of tests invented by the AI companies themselves in many cases that have obscure acronym names. So we just see these charts where they'll say, "Look, we got a 20% boost on this benchmark. We're now moving to be comparable to this machine on this benchmark."

[00:07:16]So we're not seeing major revolutionary leaps in functionality anymore. Instead, we're seeing slow and steady progress.

[00:07:23]To give you an example of like what this is like, I'm going to load up here on the screen uh a Reddit thread that's talking about the most recent release from Anthropic, which is Claude's Opus 4.7.

[00:07:34]There's a summary here of a a a vigorous conversation that happened on this Reddit thread about the new Opus 4.7. Here's the summary. The verdict is in, and it's not pretty. The overwhelming consensus in this thread is that Opus 4.7 is a massive regression and a serious downgrade from 4.6. Six users across the board are reporting a dumber, lazier, and less reliable model that feels like a step back to early chat GPT. All right, so we're not talking about this our newest model and we're not seeing revolutionary leaps from model to model. This is much more of a jagged type of frontier. Make progress here, take steps back elsewhere. Uh, another model that was released recently was OpenAI's GPT 5.5. And people seem to like this better than Opus 4.7. Um, but what like what is the magnitude of these improvements? Well, I loaded up a review. I'll put on the screen here. Um, Matt Schumer had a long review he posted online uh where he said, "A big upgrade that doesn't always feel like one." Um, I'll read you a couple things from this review. Uh Schumer is excited about continued improvement in the LLM playing nicely with coding harnesses. So um doing producing long-term plans for coding. Here's what he wrote. For serious software work, it is exceptional, thoughtful, careful, able to make many of the same decisions I would make and very good at iterating against a goal until a thing actually works. But as he also goes on to point out, um these models have been getting slowly and surely better at this type of thing uh steadily now for about a year.

[00:09:06]So, you might not actually notice much of an improvement because they're already pretty good at that. Um, here's a summary of what Matt called the biggest story about GPT5. GBT5 rounds out the weaker parts of the GPT line, design from existing context, iOS native, Mac app, security, etc. Right?

[00:09:23]So, what are we hearing here? This is what we're hearing about these latest newest models is uh slow and steady.

[00:09:30]It's like normal software updates. We improve the native Mac iOS thing. we we tweaked this functionality and now we're getting better scores on this particular benchmark. Sometimes they swing in a miss like Opus 47. Oh, the tweaks we made actually made this worse and so uh people are going to go back to the previous. There's nothing bad about that. This is just like the normal pace of software improvements. But the problem is to get from a place where we are now where almost no knowledge work is fully automatable, no major task is fully automatable to a place where all knowledge work tasks are fully automatable. We're not going to get there in another year with these slow and steady improvements. One step forward, one step back, two steps forward, one step back. Uh if we tweak this, we improve that, we improve that.

[00:10:13]That's nowhere near a fast enough pace to get us where Sulleman said we're going to get. Now, at this point, you might be saying, "Yeah, but what about coding agents?"

[00:10:22]coding agents feel like this example of a major knowledge work task software engineering that we weren't using AI heavily except for in like u more autocomplete ways and then suddenly it seemed last fall into the new year everyone was using AI coding tools within software engineering firms it's not everyone but it's like massive percentages now the way that the general public thinks about this is like well I don't know like AI quotes around it keeps getting better and better and it got good enough to automate that all of the sudden And as it continues to get better, maybe it will all of a sudden unlock other types of tasks that we can fully automate. But that doesn't really describe what happened with the rise of coding agents within enterprise software development teams.

[00:11:06]What you have to understand is that that leap was actually much as much about if not more about the quote unquote coding harnesses as it was the underlying models. The coding harnesses is the the software program written by people. It's not machine learning. It's not AI that actually calls the LLMs for ideas and then executes the LM LLM's ideas on behalf of the LLM. What really happened is there's a multi-year process of multiple companies working on these coding harnesses to make them more and more relevant. They're trying to figure out how can we get LLM's ability to produce computer code, which we've known since 2022. How do we get this ability better integrated into how actual professional software development teams operate on really large code bases? And so most of the innovation actually happened in that coding harness. They figured out all these different rules and approaches. We're going to use skill files. We're going to have a sort of simulation of memory over these skills files. Um, there's a ton of just old-fashioned like 1950s style AI in these coding harnesses like regax pattern matching using uh existing software tools to verify code. Um, so there's a lot of clues together stuff you can read. I think Gary Marcus had a good piece about this a few weeks ago because they leaked the code for the cloud code coding harness so we know what's in there. So there's just slow and steady work on building the coding harness until they could finally figure out how to make it play nicely enough with enterprise coding that we could now bring LLMs into that process. Now they also tuned the models to play better with the harnesses as well, but I think it's in the harness development that we saw that's what made AI coding relevant to uh big software development teams. So what this tells us if you want to have a similar jump in another type of major knowledge work task somewhere you have to have a lot of people iterating for maybe a year or two to try to figure out the right harness to connect properly into that particular type of job. And if we want Sulleman's claim to be true that all major knowledge work task will be fully automatable. You would need like a thousand of these teams each of them focused on another major knowledge work task trying to build a custom coding harness that works just for that task.

[00:13:17]Well they're not doing that. They don't have enough people to do that. The market isn't there and it takes experts. The one thing that people working at AI companies are experts at is software development because that's what they do. So this was like the ideal place to do this. So I think the real lesson of the quote unquote sudden emergence of coding agents is that it's actually really hard and takes a lot of focused work to try to integrate AI into individual uh types of workflows. It's not something that just happens as the models get smarter.

[00:13:47]So again, unless these companies are hiding thousands of teams working on all these different areas of knowledge work trying to find ways to subtly integrate AI into them, I do not see how we're going to suddenly have coding agent style automation in many other tasks within a roughly 12 month period.

[00:14:05]All right, my third and final point while why Sulleman is likely wrong, the functionality of LLMs are limited.

[00:14:15]So let's we're moving down this technical stack here. Let's briefly open the the black box around LLM to better understand what they can and more importantly cannot likely do no matter how big we scale them. So let's start from the bottom. We've done this before, so I'll go quick. What does an LLM actually do? It predicts tokens. You give it text and it outputs what token should come next. It has been trained to assume the text that's given as input is a real text written by a human that actually exists. And so there is a right answer for what the next token is and it's trying to get the right answer.

[00:14:51]That's at the base what an LLM does.

[00:14:54]Okay. So then how do we get long text out of an LLM? Well, you have to actually put a program on top of the LLM to keep calling it again and again what's known as auto reggressively. So you have some text, you give it as input, it outputs a token, you put that token at the end of the input text, and now you take that slightly longer input text and put it through the LLM, and you get another token. Now you take that even longer input text and you put it through the LLM, and you keep doing that until it's grown out a whole answer, and then you can return that answer to whoever made the original prompt. So what this auto reggression of token guessing gives you is basically a story completer. Here's some text and the LLM is implicitly trying to finish the story it was given as input um as accurately as possible based on all the types of text it's seen so far.

[00:15:41]Now in the original big LLMs like GPT3 it would create reasonable stories but they would be all over the place. So then they figured out how to then what's known as post- train or tune these LLMs of the possible ways they could go to tune them towards certain categories of types of text and away from others. So GPT35 which was the version that powered the original chat GPT was tuned to uh to be uh to think about the stories it's completing as answers to questions. So it put it into an uh you're given a your prompt the input's a question you're trying to answer the question. Um, and now it's much easier for like average users to deal with. All right, so they're story completers. Now, we don't want to downgrade that, right? This this isn't a dismissal like ah it's autocomplete because what we discovered particularly with GPT4 is that as we scaled these things up in order to successfully complete stories in reasonable ways like these LLMs actually encoded a lot of really interesting rules and logic, right? If my story involves a math problem, to complete that story in the right way, the LLM is going to have to have some math logic built in there because otherwise the story is not going to seem reasonable.

[00:16:50]And this was really the big discovery of GPT4 was like, wow, there's all of these abilities that were implicitly coded into this LLM during its training. It's just trying to complete the stories, but we trained it on so much stuff for so long that it learned all of these types of abilities that allows it to win the story, you know, the complete stories.

[00:17:08]Well, like it it understands humor. It understands math. It understands computer code. It understands basic logic. If it's seen a game enough time, it has some sense of like generally what the game rules are, what a valid move in that game looks like. It's pretty it was pretty incredible um what GPG4 uh could actually do. And there was this idea of like if we kept scaling these things bigger and bigger. Yeah, sure they're story completers, but um they can eventually have so many rules and logics programmed into them to win this game that like they will actually have like a human level intelligence and then we can just ask the use the LMS brain to automate everything. That was the vision. Okay. Then what happened is we learned by the summer of 2024 is that this type of just scaling was hitting a wall. just trying to make these models bigger and train them longer was not appreciably leading to new functionality being encoded into them like we had seen before.

[00:18:00]And this picked up a whole era that really started with the alphabet suit models out of OpenAI in the fall of 2024. It kicked off a whole era of post-training and tuning where okay they get further appearances of improvement in these models. We can't just scale them bigger. Um, we can do things called post-training where if we have very specific data sets where we have questions and exact right answers, we can we can tune an already pre-trained model to be better at uh using its already wired intelligence to answer this type of problem. And that's why we saw starting 2024 a big focus on the type of things for which we had data to do this tuning, reasoning, math, and computer coding. It's why why weren't they talking about other types of things you'd want an LM to do because it would be economically very valuable because all we can do now is tune these and start and get sort of like steady improvements in areas where we happen to have a lot of uh highly structured data to do the tuning on it. And that's really where we've been since 2024. And that's why we have more of this um slow but steady improvement with LLMs on particular benchmarks and no longer these big general leaps like we had during the pre-scaling era when we saw like 3 to 35 to four. Okay, so this is a problem uh for automating work because uh that we know we can't scale these LLMs just to have like a human level of reasoning on things and for most things we do in knowledge work most of the skilled things we do we don't have highly structured data sets like we have for computer programming or math so we can't even tune the pre-trained LLMs to be better at things that a lot of people do in their jobs so this is sort of a wall that's hard to get past when you leave uh where we are right now with LLMs. Now, you might say, "Yeah, but what about workplace agents?"

[00:19:45]A lot of what people do in knowledge work is actually not that skilled. It's individual things an LLM can do. Send an email, uh get information about an upcoming conference, move information to a spreadsheet, build slides and send it to the team. Like, actually, a lot of things we do in knowledge work is what I call shallow work. It's things that don't require a particularly high level of skill. So why can't we have at the very least something like a knowledge workplace administrative assistant that can automate a lot of boring stuff we do in knowledge work like we have in computer programming where where coding agents can do uh multiple steps of work on behalf of the programmer. So at the very least this wouldn't be sulie man's prediction but it would get us closer.

[00:20:25]Why don't we have those either? Well it turns out that uh building these type of building these type of tools is difficult. I wrote an article about this back in January for the New Yorker. It was called, "Why didn't AI transform our lives in 2025?" And in that article, I looked at the question of why don't we have more workplace agents? We were told at the beginning of 2025 that we would this this would be the year of the agent, not just coding agents, but for all the things you do in work, you would more people would have these AI assistants.

[00:20:59]Why did that largely not happen? Well, when you actually study how these agents work, the problem is the plans, the multi-step plans that agents execute are the result of LLM prompts. You have a harness, which is a computer program written by a person that prompts an LLM and says, "Give me a multi-step plan for doing X." And then it takes that text and it goes step by step to harness and executes those things um on behalf of the user. for computer programming.

[00:21:25]They've they've learned how to be pretty good at making those multi-step plans because the the the option space of what you do when you're building a computer program is very narrow and what they're doing is very verifiable. And so actually the the harnesses for coding agents will call nonAI tools. They can verify things. All right, this step has a clear success category like we the code compiles and passes these tests and we're going to run another program that's going to check the code, make sure it compiles and passes those tests.

[00:21:52]And if it doesn't, we can go back to the LLM and say, "That plan didn't work.

[00:21:54]Give us another one." Right? Um, it's very well set up for doing this. But once you're in the more, so I wrote about in my New Yorker article, once you're in this more general world of more ambiguous sort of knowledge work task, what you're going to get out of the LLM is a reasonable sounding plan.

[00:22:09]Reasonable sound because that's what it does. It's a story completer. Reasonable sounding plans get you in trouble. You need correct plans. And the way that like humans actually make plans is we do a few things. One, we test a bunch of possibilities internally to see what makes the most sense. And two, we have some sort of notion of correctness and a world model we can use to evaluate plans to see, does this actually do what I want it to do? Are there any mistakes along the way? That's not how LLMs operate. Again, we're just auto reggressively producing tokens. So, you get a reasonable sounding plan, but it's not stepping back. It doesn't have a world model to test it against. It can't have hard-coded rules that it consistently applies. It has no ability to do future simulation of possible outcomes and so we just get a story that sounds like a good plan but often has issues along the way and if an agent is just going to automatically execute these things we get into trouble.

[00:23:01]All right. So making agents even just like administrative assistants in non-computer programming areas is also a very hard problem uh and not one that we have a good solution to. Um, we also underestimate the degree to which if you watch a programmer using a coding agent, they're constantly tweaking and reasking and adjusting. There's a huge amount of work to try to get consistently useful uh output out of these agents. And most knowledge workers just simply aren't going to do that or have the technical chops to pull that off. Earlier this year, actually in the fall, OpenAI even mentioned they were slowing down or reducing their non-coding agent projects because they weren't working and they want to focus on chatbt and their coding agents. All right, so if we put all these things together, we put these three reasons together, I think it's clear that Mustafa Sulleman's claim that AI would fully automate most knowledge work jobs by next spring really is an outlier opinion.

[00:23:53]is an opinion that is count contradicted by the statements of other CEOs, the reality of the rate of progress of LLMs over the last couple of years, which is not nearly fast enough to get where he says we're going to get, and there the technical limitations of these models.

[00:24:06]They're just LLMs with a harness on it, is just not a great setup for automating a lot of different types of knowledge work jobs.

[00:24:14]So, why then might Sulleman be pushing this story? What's in it for him? I'm going to play you a quick clip of one potential explanation. It comes from the absolutely agentic channel. Let's hear what they had to say there.

[00:24:27]>> And it's worth noting that this interview comes at a particularly interesting moment for Microsoft.

[00:24:31][music] Their stock recently took a hit as markets questioned whether the enormous capital [music] expenditure on AI infrastructure is going to pay off. So, Sullyon has a motive to sell the dream.

[00:24:44]More generally, claims about AI massively disrupting the job market, like we just sort of heard there, tend to be very useful for the major AI companies. They know they can just say wild things about economic disruptions and they're not going to be challenged on these predictions. Even if they don't come to be true because it matches a general vibe of there's disruption and we that's exciting or scary, we want people to to be worried or whatever it is, they're not going to be challenged, so why not? And it makes them seem like they're working on the most important technology ever, which is exactly the type of technology if you're an investor, you would say, "Sure, anthropic, we've given you $60 billion and you've made $5 billion to date of revenue, but like, hey, I'm not going to do normal expectation accounting with you because you're building the most important technology ever. And so none of these numbers matter. You're going to automate the whole economy." So, it's super in the favor of the CEOs to say these things.

[00:25:36]Now to be clear, I don't want this to make it seem like LLM based tools are irrelevant to the workplace. They're not. As I promised, I wanted to talk briefly about the ways in which LLM and LM based tools are actually useful in non-coding knowledge work sectors.

[00:25:51]There's five things I want to quickly mention that right now we're seeing or will soon see useful applications of LLM based tools. one, sifting through reasonable amounts of text to generate summaries or to find useful examples.

[00:26:04]Because of the way LLMs are built with what are known as attention sub layers, they're very good at selectively turning their attention to relevant parts of the input text when they do their processing. So, if the text you're feeding to an LLM is not too expansive, it's really good at like find examples of this. Summarize the cases of this that shows up. Look for examples in here of X in this text that should catch my attention.

[00:26:29]They're really good at that and that can be useful. There's a there's a lot of like a time otherwise time inensive things that humans would have to do um that this makes much. You could do this with natural language prompting. If the text gets too long, the accuracy is going to fall and then you're going to have to do uh something more advanced, which I'll talk about in a second. The second major use we're seeing of these in the knowledge workplace is data formatting. It's very good again at summarizing text and rewriting it in different formats, right? So, if you can say, hey, take these 10 consumer comments and go through and make a bulletoint list or put it into a slide of like the five most common things they say. It might not be that precise, but I'll do a good job of reformatting text or clean up this data so that I can put it inside of a spreadsheet. Now, again, if the amount of data gets too big or your precision level, like it needs to do this exactly right every time, gets too high, just prompting an LLM is not going to necessarily be accurate enough.

[00:27:24]But that brings us to the third use, which is emerging more and more. For highly technical users, you can use coding agents to produce small programs to do this type of processing on more precisely on bigger amounts of data. So if you give me like 10,000 rows of a spreadsheet and say go through here and clean it up in this way, you can't just feed that to an LLM, the context is too big. It's going to produce reasonable looking stuff with lots of mistakes. But what you can do if you're technical enough is have something like cloud code produce a quick Python script that can read through the file and very systematically and correctly make the changes and then it can once you have that program you can feed it the biggest possible file. And I think that's useful for people who are more technical within knowledge work. A lot of people are using it as a better Google. A lot of what happens with these chatbot now is basically doing a Google search and then feeding that to an LLM which can then summarize it for you. And that's very useful for a lot of people. It can summarize data from a Google searches effectively or put them in useful formats. I think that is very useful.

[00:28:25]And I'm hoping soon we're going to have much better calendar or appointment management. That's one agentic thing that's very narrow and very good for an LLM to do. Find an appointment with these rough roughly described parameters, you know, on my calendar. LM are very good at that. Uh I wrote an article last year about email sorting with LLMs and they're getting that's another good use. I mean, it's expensive from a token perspective, but hey, is this email you can give it like very natural language rules for what you care about and what you don't know how to filter it. And LM can act on that and act on an email and filter it. So, I think there's these uh focus sort of filtering and agentic interactions in the workplace where LM would be very useful. So, these are all really cool tools.

[00:29:06]There's a couple other things people are doing right now in the workplace, which I think they shouldn't. I don't think they should be writing so much with LLMs. If an LM is producing your slide deck or it's producing the emails you're sending, then you shouldn't be using that slide deck and you shouldn't be sending those emails. The information content was too low. You should have have something much simpler. Um, I'm also not a big fan of using LLM to quote unquote refine your thinking. They're syncopantic, hallucinatory, and emotionally manipulative is not a good way to try to sharpen your thinking. You need to read hard things. You need to write to organize your thoughts. You need to talk to other real people. But there are real good uses of LLM in the workplace that are expanding. But none of this is a future in which by next spring all the jobs are automatable.

[00:29:44]That is just not realistic.

[00:29:46]All right, to conclude, I promised you earlier in this episode that I uncovered a quote unquote conspiracy while I was working on this story, um, which I thought was fun. So, I will tell you about it. All right, so the Sulleman interview with the Financial Times went up in February. Uh, I I'll load it on the screen like, so here it is. Um, Mustafa Sullean sets out Microsoft's AI goal for humanist super intelligence.

[00:30:11]Okay. February 12th when that interview went out, there were dozens of articles about the part in that interview that I played at the beginning of the show where he said 12 to 18 months all of these jobs where people sitting in front of a computer will be fully automatable by AI. Uh it was clipped dozens of times all over YouTube and social media. You can see clips of that part of the interview. It was written about in many publications, you know, where they they they link to the interview and have the actual exact quote, but if you go to this official video today on the Financial Times website, they've edited it out, so it's gone. He leads right up to the point where he said that quote and then there's an awkward cut where it cuts from a close shot to a wide shot and he's jumps to another topic that kind of you you notice it when you first listen to it. like that's a weird jump. Why was he going from this to that? So, they edited out of the original video. Now, it was too late because again, dozens of people already clipped it and spread it and major publications wrote about it.

[00:31:15]So, I don't know why they thought they'd get away with editing it out, but it is now actually gone from the official version of the video. Now, I don't know why they did this. Here would be my best guess. I think Sullean saw these other CEOs making big bombastic claims and getting a lot of attention and hype for their products and they're like we got to get in this game too and so he tried it but he went too far.

[00:31:42]That's what I think happened. He's like I can do like an Amadeday here. Uh what if uh all jobs will be gone in a year and he went too far and I think after the fact whoever it was on the executive team or the lawyers or whatever were like this is not we don't want to be saying these type of things. This is way too drastic. It makes us seem a little bit dirty. So like, all right, go edit it. Go edit it. So they went back and they edited out of the video. That's what I I don't know if that's what happened. That's what I think happened, but it's just too late because it was already spread out there. But if that's true, then here's what I'm going to say.

[00:32:13]Um, look, I don't know. Actually, here's what I'll say. I don't know if that's true or not, but I'm going to take a page out of other AI commentators and say, you know what, that matches my vibe about what's really going on. I think the idea that they edit it out because they got it wrong is directionally true.

[00:32:28]So, in the spirit of all the other AI reporting going on out there, I'm just going to claim that's true because it feels right to me. All right, that's it for this week. Um, we'll be back next Thursday with another AI reality check episode. Also, stay tuned on Monday on this feed. We have uh advice episodes which I think you'll like. So, you should check those out as well. They provide practical advice for seeking depth in a distracted world. It's all good. So, stick around. Until next time, remember, you should care about AI, but not everything that's said about it.