Source Details

{
    "source_type": "extension",
    "content_hash": "67466f21c77566b68947ef59adb875a276d44d56b02126b5e82ac2316c8adc94",
    "submitted_via": "chrome_extension",
    "extension_version": "1.0.18",
    "parsed_content": "\u201cI can give it tasks and just walk away.\u201d\u201cIt captures details I would have otherwise forgotten.\u201d\u201cThere\u2019s so much low-hanging fruit.\u201d\u201cThe important thing is to maintain a reserve of skepticism.\u201d A burst of experimentation followed ChatGPT's release to the public in late 2022. Now many people are integrating the newest models and custom systems into what they do all day: their work.Chefs are using A.I. to invent recipes; doctors are using it to read M.R.I. and CT scans; scientists are unlocking discoveries. It\u2019s helping workers with their day-to-day tasks: writing code, summarizing emails, creating ideas, generating curricula \u2014 even as it still makes plenty of mistakes.Recent surveys have found that almost one in five U.S. workers say they use it at least semi-regularly for work. Twenty-one people told us how.People are using A.I. to \u2026Select wines for restaurant menus Sam McNulty Restaurant owner and operator Mr. McNulty, who owns restaurants, brewpubs and dance clubs in Cleveland, uses ChatGPT to analyze sales reports and brainstorm how to grow sales. He\u2019s also used it to help pick wines. He sent a \u201cvoluminous\u201d wine portfolio from a distributor to the chatbot and gave it some instructions \u2014 specific pricing and particular regions among them \u2014 and got back a list, including:Herdade do Espor\u00e3o Monte Velho Branco Region: Vinho Regional Alentejano, Portugal Grapes: Ant\u00e3o Vaz, Roupeiro & Perrum Wholesale Est.: $7-9 per 750 ml Why It\u2019s Great: A crowd\u2010pleasing white that combines citrus, stone fruit and saline notes with bright acidity \u2014 an ideal food\u2010friendly pour for small plates or seafood.\u201cThe results were astonishingly good and saved me and my team countless hours of meetings with wine reps, tastings and debate,\u201d he said. \u201cThe only part of the wine-program building process I missed was the tastings ... so far the A.I. can't recreate the joy of taking that sip.\u201dDigitize a herbarium Jordan Teisher Curator and director There are eight million dried plant specimens at the Missouri Botanical Garden herbarium in St. Louis. Now A.I. is helping identify them.Experienced taxonomists can quickly recognize most specimens, said Mr. Teisher, but that requires years of training.So the garden is building an A.I. model using spectral data \u2014 the pattern of light reflected by the plant. Leaves from many different kinds of plants are scanned, labeled and put into the model as training data. Then new plants can go through the same process, and the model will identify them. If the model is quite certain that the spectral data look the same as they do for other plants, it\u2019ll say so. If not, the plant can go to an expert. Leaves are placed on a black plate to measure their \u201creflectance spectra,\u201d part of building an A.I. model that can identify new specimens. Nathan Kwarta, Missouri Botanical Garden \u201cWe can cut down on the time expert taxonomists are spending on common species,\u201d Mr. Teisher said. \u201cRather than them getting five boxes of plants that come in, they can get a small box that says, \u2018Here are the ones the model isn\u2019t sure about.\u2019\u201dThose could be species so rare or infrequently seen in the herbarium that the model simply couldn\u2019t match it, or a new species entirely.And this kind of project is only possible, the garden staff said, because of advances in cheap computing power. The GPUs necessary to train A.I. models quickly are easier to get than before. And the garden has enough funding to process several hundred thousand specimens. Identified specimens can be used for a variety of research, including on biodiversity and climate change.\u201cWe want these data to be compatible with other institutions, so we\u2019re collaborating closely,\u201d Mr. Teisher said. \u201cWe have the money to do a big chunk of our herbarium, but ultimately we want this to be a tool usable by everyone.\u201dMake everything look better Dan Frazier Designer and small business owner Mr. Frazier designs and sells things like bumper stickers and magnetic signs. To help with the graphic design, he uses Adobe Photoshop\u2019s Generative Fill, a two-year-old A.I. feature that adjusts images automatically.\u201cIf I take a picture of a product, and don't like the glare or reflection I see on some shiny surface, I can use generative fill to \u2018imagine\u2019 that part of the photo, and usually one of the resulting images will be acceptable to me,\u201d he said. \u201cOr if I want to use a head shot of a politician on a bumper sticker, and I want to show a little more of the coat or shirt than appeared in the photo I am using, I can use generative fill to imagine that additional clothing.\u201dA problem that might have taken 20 minutes to address now takes 20 seconds, he estimated.In one recent case, he wanted to post an image of a bike helmet he\u2019d built himself, but he didn\u2019t think he was the best model for it. So he used Photoshop\u2019s A.I. to generate a woman\u2019s face. \n Dan Frazier But A.I. can only fill in the gaps for Mr. Frazier. \u201cI have found generative fill to be less useful at creating images from scratch,\u201d he said. \u201cI once wanted to create an image of Joe Biden looking like one of the founding fathers, maybe like George Washington. But I was not happy with any of the results I was getting. I ended up melding a photo of Biden with a lifelike painting of Washington using traditional Photoshop techniques.\u201dCreate lesson plans that meet educational standards Manuel Soto E.S.L. teacher Mr. Soto, an E.S.L. teacher in Puerto Rico, said the administrative part of his job can be time consuming: writing lesson plans, following curriculum sent forth by the Puerto Rico Department of Education, making sure it all aligns with standards and expectations. Prompts like this to ChatGPT help cut his prep time in half:Create a 5 day lesson plan based on unit 9.1 based off Puerto Rico Core standards. Include lesson objectives, standards and expectations for each day. I need an opening, development with differentiated instruction, closing and exit ticket.After integrating the A.I. results, his detailed lesson plans for the week looked like this: English as a Second Language sample lesson plan. Manuel Soto But he\u2019s noticing more students using A.I. and not relying \u201con their inner voice.\u201dInstead of fighting it, he\u2019s planning to incorporate A.I. into his curriculum next year. \u201cSo they realize it can be used practically with fundamental reading and writing skills they should possess,\u201d he said.Make a bibliography Karen de Bruin Professor of French and scholar of 18th-century French literature Anyone who has ever assembled a \u201cworks cited\u201d section knows the dizzying array of styles, formats and specific punctuation rules required for a bibliography. What\u2019s the Chicago Style rule on how to cite books? Do you use quotation marks or underline? Or is it italic? What about in A.P.A. style? A.I. has freed Ms. de Bruin from the most annoying parts of the task. \u201cNo more consulting handbooks, guidebooks, cheat sheets, Purdue Owl, fretting about the right punctuation, whether guidelines have changed, and how to cite a three-volume work written in the 18th century, translated by God knows who, edited by Jesus only knows, and originally published where?\u201dAll of this has been replaced, in her words, by \u201cpeace, serenity and Claude\u201d (the large language model).She uses prompts like:Please cite in MLA format the book University Finances by Dean O. Smith.orGive the mla citation for this article: www.chronicle.com\/article\/higher-eds-financial-roller-coasterOccasionally Claude cites \u201cDoe, Jane,\u201d and Ms. de Bruin challenges the answer.\u201cThen, only then, does it respond that it took its best guess at the author because the article was behind a paywall,\u201d she said.Write up therapy plans Alissa Swank Psychotherapist Ms. Swank uses A.I. to take unstructured notes from a visit and turn them into SOAP notes, which is a structured documentation format for health care providers. (It stands for Subjective, Objective, Assessment and Plan \u2014 a way of summarizing the visit and the next steps.) It saves her a couple of hours each week, she estimates, \u201cbut more so it helps me complete the task that is so easy to put off.\u201dAs a \u2018muse\u2019 Marya Triandafellos Visual artist Ms. Triandafellos uses A.I. as inspiration for her art practice. She uploads dozens of images of her artwork to get the A.I. model to understand her style, then guides the model with prompts to generate new works based on her style. What she gets back are hundreds of abstract images in a grid: Marya Triandafellos She studies them the way a psychiatric patient interprets an inkblot test.\u201cI looked at each image and wondered what it reminded me of, reaching my subconscious,\u201d she said.From there, she sorts them into themes and uses them as a base for a more fully finished work. She also asks the model to be her critic:Please act as an art critic and evaluate this piece based on its subject, themes, how it makes you feel, and historic connections. Consider how it may be connected to science or math. Then, provide me with an appropriate title.\u201cIt may not be as nuanced as a human art critic,\u201d she said, \u201cbut it does decipher key aspects of the work which I refine further.\u201dShe doesn\u2019t use A.I. to create final pieces, though: \u201cI tried \u2014 and was bored and frustrated.\u201dDetect leaks in a water system Tim J. Sutherns Company president When a water system springs a leak, you might not notice until it becomes a big problem. Mr. Sutherns\u2019s company, Digital Water Solutions, is trying to catch leaks early by placing small sensors inside fire hydrants that record the noise water makes as it flows through the pipes. That data is fed to a machine learning model that looks for certain patterns suggesting a leak.It\u2019s a relatively simple concept, but it\u2019s hard to reproduce quickly, said Mr. Sutherns, in large part because every system is different: different pipe material, sizes and pressures.\u201cIf we had to build individual machine learning models for every one of these unique systems, it would take us months, a whole bunch of data scientists,\u201d he said.Instead, the team uses \u201cautonomous machine learning.\u201d A.I. figures out, on the fly, what the parameters of the model should be for a specific system, meaning the company doesn\u2019t need to know anything about the system ahead of time \u2014 it just has to start collecting data. Within a couple of weeks, typically, the models can provide some information on possible existing leaks. Digital Water Solutions Mr. Sutherns started the company in 2018, but recent advancements in machine learning, cheaper computing power and data storage have made the business far more feasible.Small water systems, serving fewer than 10,000 people, make up the vast majority of water systems in the U.S., and have small budgets. Offering the technology to those systems at a reasonable price? That wouldn\u2019t have been possible a few years ago, he said.Just write code Chris O\u2019Sullivan Chief technology officer and company co-founder It\u2019s one of A.I.\u2019s simplest and most common use cases \u2014 one that even the A.I. engineers are leaning on: writing code. Mr. O\u2019Sullivan is one of them: As the C.T.O. of DraftPilot, a legal A.I. company that helps lawyers with contract review, he frequently uses Anthropic\u2019s Claude Code.\u201cI can give it tasks and just walk away,\u201d he said. \u201cIt writes the code itself.\u201d Chris O\u2019Sullivan Type up medical notes Matteo Valenti Primary care physician At Dr. Valenti\u2019s hospital, an A.I. tool, Abridge, is built into the electronic medical record system to take notes when he meets with patients. The tool listens to his conversation with the patient, then creates an organized record of the visit \u2014 the kind he would otherwise have to produce manually. Abridge It saves him about an hour each day, he estimates, \u201cbut the real benefit is that it captures details I would have otherwise forgotten.\u201d If a patient comes in for diabetes, but briefly mentions back pain, that aside makes it into the record whether or not he remembers it. And he\u2019s able to focus on having a real conversation with patients, without transcribing every word.He worries that the tool may replace human scribes. But for providers on tight budgets, it makes a difference. \u201cFor those of us in primary care who are drowning in paperwork,\u201d he said, \u201cthis will be a plus.\u201dRun experiments to figure out how the brain encodes language Adam Morgan Postdoctoral fellow For his research in cognitive neuroscience, Mr. Morgan works with neurosurgery patients. While their brains are exposed, he runs experiments that attempt to examine how the brain encodes things like language and meaning \u2014 often by asking them questions while directly measuring their neural activity.Because there\u2019s usually limited time and subjects on whom he can run experiments, he has to prioritize research topics. That\u2019s where A.I. comes in.Like a human brain, artificial neural networks take some kind of input (words, say) and produce outputs (other words). For the human brain, what happens in the middle is something of a black box, but we know that words we hear are translated into neural activity that represents meaning, then decoded into other words. Mr. Morgan says artificial neural networks do something similar, only using numbers.\u201cThere\u2019s good, and growing, evidence that L.L.M.s encode syntax and words in a similar way as the brain,\u201d Mr. Morgan said.But unlike with a brain, you can directly examine these encoding processes in a large language model just by looking at the code. So the A.I. can act as a pseudo brain to test hypotheses about language that are hard to test in real brains.\u201cIn my work, I figure that if I find that the middle layers of a computer model are sensitive to a particular property that I'm interested in in the brain, it's a decent indication that the brain might care about that,\u201d he said. Help get pets adopted Kristen Hassen C.E.O. Ms. Hassen\u2019s company, Outcomes for Pets Consulting, works with large animal shelters to decrease euthanasia rates and shorten animal stays. She uses A.I. to come up with ideas:Give me 50 ideas for adoption promotions focused on senior pets who have lost their homesOne of them was:Lifetime of Love: Side-by-side then and now photos of pets who have lost their longtime families and a call to give them love again.\u201cWe\u2019re definitely going to do that one,\u201d she said.Check legal documents in a D.A.\u2019s office Chris Handley Director of operations and chief of innovation Mr. Handley works in the Harris County District Attorney\u2019s office in Houston, the third-largest jurisdiction in the country. He recently built a custom large language model that helps prosecutors and the police avoid errors when filing arrest paperwork.After booking someone, the police type up their account of events, and that report goes to the D.A.\u2019s office. It then goes straight into Handley\u2019s L.L.M., which does a series of checks, looking for issues a judge might later catch \u2014 a typo, a missing piece of information about the arrest, a slightly incorrect charge, a full name of a sexual assault victim rather than initials, all of which could and do slow the process.\u201cWhen people think of A.I., they think of chatbots, or they think of Skynet, facial recognition,\u201d Mr. Handley said. \u201cWe're not doing any of that. For us, there's so much low-hanging fruit. Just making sure our paperwork doesn't have mistakes on it.\u201dThey\u2019ve been testing the program and working on a larger rollout. A colleague tried it and said it reduced her work time by 50 percent. Mr. Handley now wants to pilot a model that could work with police officers while they\u2019re first filing charges from the scene.But the models are not useful for everything yet. He trained one model on case law and asked it about one of his cases.\u201cIt very confidently went on and on about these made-up facts that had nothing to do with my case,\u201d he said. He deleted the model.Get the busywork done Sara Greenleaf Project coordinator Ms. Greenleaf works for a health insurance consultant, and many of her duties are administrative: drafting contract documents, scheduling meetings, editing PowerPoint slides, signing people up for conferences, and so on.She turns to ChatGPT to get all those tasks checked off. It helps her summarize \u201caction items\u201d from a long chain of emails; proofread her emails; create contract templates; search through long documents like benefit summaries; and compare documents when she suspects there might be small differences.But it wouldn\u2019t help her with her first career: pianist.\u201cIf I hadn\u2019t had this experience of working in an office, I think I\u2019d be mostly horrified by A.I.,\u201d she said. \u201cI never use it in my creative life, and am very worried about its implications for the arts.\u201dAnd it hallucinates sometimes, she added, so she checks and cross-references her results carefully. \u201cA.I. is not doing my work for me,\u201d she said. \u201cMost of the time it\u2019s just getting me started with a task or prompting me to think of something in a different way.\u201dReview medical literature Michael Boss Medical imaging scientist Mr. Boss oversees the use of M.R.I., CT and other scans in clinical trials, ensuring that imaging is done to protocol and working on standardization efforts. He\u2019s reading medical literature nearly every day \u2014 and he uses ChatGPT, Perplexity, Undermind and more tools for that.That means he can say something like:Identify relevant imaging biomarkers and their reproducibility as evidenced by ICC, CCC, or wCV in primary prostate cancer as used in interventional studies.And get back a result like: He doesn\u2019t rely much on A.I. summaries; instead, the chatbot\u2019s response gives him a sense of what scientific literature might be relevant to his question and worth reading in full.\u201cUsing A.I. has profoundly sped up the process,\u201d he said.He\u2019s learned to be very careful about chatbot summaries in particular. Recently he asked ChatGPT a question about M.R.I. diffusion, an area where he\u2019s made some contributions. The response misattributed his work to a person who appeared not to exist \u2014 frustrating for a scientist whose reputation is built on credit, and alarming for a chatbot user.\u201cI find that ChatGPT's current approach is very much a groupthink summary, if you take it at face value,\u201d he said. \u201cThat is potentially dangerous. However, taking its results with skepticism, you can use the results to seed additional searches, or additional prompting to get to the right answer.\u201dPick a needle and thread Nicole Goldman Fiber artist For Ms. Goldman\u2019s work as a fiber artist, she often needs to know the best stabilizer to use, or the best glue, for a particular project.\u201cI've used Claude to resource materials, to help me decide what size needle and thread I should be using for a particular project, to give me technical information,\u201d she said. \u201cWhere I might have \u2018Googled\u2019 before and had to sort out a huge variety of information and sources, this definitely cuts right to the chase and organizes the information so much more quickly and succinctly.\u201dRecently she asked Claude for a didgeridoo pattern. The final product ended up more like a bird, she said, but she didn\u2019t mind \u2014 she considered it a collaboration with the A.I. Nicole Goldman (More politely) let band students know they didn\u2019t make the cut Deb Schaaf Music teacher and jazz director Ms. Schaaf is a music teacher in a competitive high school jazz program. Not everyone can make the cut, and she has to deliver the news. She uses A.I. to help let down her students firmly but gently.\u201cI discovered my favorite prompt after asking the A.I. for more diplomatic language in a message about the need to fire a drummer,\u201d she said.Her initial attempts were \u201cso padded with feel-good fluff that it became nearly twice as long and obscured most of the issues.\u201dAfter some back and forth, she finally landed on a prompt that worked:Make it more Gen XThe results were what she was hoping for, \u201ca much more direct message that was thoughtful, but didn\u2019t sound like Mr. Rogers on molly.\u201dHelp humans answer more calls at a call center Thor Dunn Chief, Customer Service Center California\u2019s Department of Tax and Fee Administration is responsible for tens of billions in state revenue each year. And because taxes are complicated, its main call center gets hundreds of thousands of calls a year. That\u2019s where the department thinks A.I. can help. It\u2019s testing a system using a version of Claude trained on state data.During a customer service call, the A.I. reads a live transcript and suggests an answer. The human agent on the call can then click through to the reference material linked in the A.I.\u2019s answer, and decide whether it\u2019s right. The goal is to help the real people answering calls sift through material on more than 16,000 pages of reference material on taxes and fees.Early tests showed a 1.5 percent improvement in the time it took to process calls, and Mr. Dunn thinks that could rise as call center agents become more familiar with the system. The model is working better now than it was even earlier this year, thanks to improvements in Claude.Help translate lyrics from the 17th and 18th centuries Richard Stone Orchestra co-director Mr. Stone co-directs the Philadelphia Baroque Orchestra, and as part of that job translates lyrics for renaissance and baroque vocal works. He has knowledge of the main singing languages \u2014 Italian, French, German and Latin \u2014 but only the way they are currently spoken and written. Versions from hundreds of years ago were different, and less standardized.\u201cThe A.I. helps me to gain the experience that my conservatory training didn't include,\u201d he said. He does all of the initial translation on his own and uses A.I. more as a \u201cconsultant or a tutor\u201d to check his work.When there\u2019s a passage he\u2019s unsure of, he\u2019ll show both the original and his translation to the A.I., going back and forth to come up with something he feels more confident about.\u201cThe important thing is to maintain a reserve of skepticism,\u201d he said. \u201cIt will make things up, so when I get suspicious I will quiz it.\u201dMr. Stone was recently trying to crack this phrase in Italian: Richard Stone via Stift Heiligenkreuz Musikarchiv The first word gave him trouble.\u201cI transcribed the Italian word \u2018pramo,\u2019\u201d he said. \u201cI invested so much energy on my own and working with the A.I. on figuring out what \u2018pramo\u2019 could possibly mean. I eventually recognized the word as \u2018bramo\u2019 (I desire\/wish). It could have been an unattested form of the word or an outright scribal error. That sort of intuitive leap is not something the platform I use is remotely good at.\u201dAnd the final translation?Bramo che sia cos\u00ec per tuo contento. I wish it to be so for your happiness.Explain my \u2018legalese\u2019 back to me Deyana Alaguli Lawyer Ms. Alaguli uses this prompt with Google Gemini to help see if her legal writing is confusing:I understand you're not a lawyer, tell me what a layman might understand from this paragraphYou can\u2019t count on A.I. to accurately interpret legal or technical jargon, she said, but it can be great for helping build your case. She also uses it to prepare for hearings and to help practice closing arguments.\u201cIt can understand your arguments, or help you anticipate holes in your case, better than a colleague can,\u201d she said. \u201cIt's faster, unbiased, not worried about hurting your feelings.\u201dDetect if students are using A.I. Matthew Moore High school English teacher Mr. Moore uses Magic School A.I. and ChatGPT to generate worksheets, rubrics, images and educational games for his various English classes. And his students are using it, too.\u201cIt does feel hypocritical to tell them not to use it when I am using it,\u201d he said. But he turns to A.I. to make sure they are using it in permitted ways.He remembers a ninth-grade student who turned in \u201ca grammatically flawless essay, more than twice as long as I assigned.\u201d\u201cI was shocked,\u201d he said. \u201cAnd more shocked when I realized that his whole essay was essentially a compare and contrast between O.J. Simpson and Nicole Brown Simpson.\u201dThat was not the assignment.\u201cThe A.I. detection software at the time told me it was A.I.-generated,\u201d he said. \u201cMy brain told me it was. It was an easy call.\u201d Matthew Moore So Mr. Moore had the student redo the assignment \u2026 by hand.But, he said, the A.I. detectors are having a harder time detecting what is written by A.I. He occasionally uploads suspicious papers to different detectors (like GPTZero and QuillBot). The tools return a percent chance that the item in question has been written by A.I., and he uses those percentages to make a more informed guess.\u201cWe are, likely, less than a year away from when teachers cannot reasonably discern between A.I. writing and student writing,\u201d he said. The more sophisticated A.I. papers can imitate the writing level of a high school student. (Some students even feed their A.I. papers into another website like Humazine A.I. to try to make the writing feel more natural.) \u201cOnce we pass that threshold, we will no longer be able to accept any typed essays or writing assignments from students. It will all have to be under testing conditions, or they will have to write it all by hand.\u201d \nHow are you using A.I.?**0 wordsIf you use a chatbot, paste in specific prompts and answers that were particularly useful or interesting. If you use special A.I. tools, tell us about them.0 wordsWhat is your email address?*What is your name?I am open to a New York Times journalist contacting me about other reporting projects.By clicking the submit button, you agree that you have read, understand and accept the Reader Submission Terms in relation to all of the content and other information you send to us (\u201cYour Content\u201d). If you do not accept these terms, do not submit any content. Of note:Your Content must not be false, defamatory, misleading or hateful, or infringe any copyright or any other third-party rights or otherwise be unlawful.We may use the contact details that you provide to verify your identity and answers to the questionnaire, as well as to contact you for further information on this story and future stories.Submit",
    "ai_headline": "How People Are Using A.I. at Work",
    "ai_simplified_title": "People Integrate AI into Work Tasks",
    "ai_excerpt": "This article explores how various professionals are integrating AI into their daily work, from chefs using AI for recipes to doctors using it to analyze scans. It highlights specific examples and the benefits and challenges of using AI tools.",
    "ai_subject_tags": [
        "Artificial Intelligence",
        "AI Applications",
        "Workplace Technology",
        "Productivity",
        "Automation",
        "Technology",
        "Business"
    ],
    "ai_context_type": "Analysis",
    "ai_context_details": {
        "tone": "informative",
        "perspective": "neutral",
        "audience": "general",
        "credibility_indicators": [
            "expert_quotes",
            "real-world examples"
        ]
    },
    "ai_source_vector": [
        0.0071444088,
        0.00023026971,
        0.0033394285,
        -0.09067639,
        0.0036422526,
        -0.016030839,
        -0.007995639,
        0.016406618,
        -0.0047583366,
        -0.004959632,
        -0.021874601,
        -0.011233428,
        0.03731074,
        0.0062449616,
        0.12959184,
        0.00063738396,
        -0.001532315,
        0.0092556365,
        0.027685456,
        -0.006472447,
        0.022193348,
        0.02419749,
        0.010707721,
        -0.009645483,
        0.006402225,
        -0.010737548,
        0.03313449,
        -0.012520157,
        0.036575396,
        0.010723435,
        -0.010947567,
        0.023039585,
        -0.00042804307,
        0.01642956,
        0.011895059,
        0.0147907715,
        0.006452902,
        -0.03273567,
        0.010742348,
        0.013628179,
        0.012418411,
        0.0030063766,
        -0.017015094,
        -0.009073815,
        0.021096647,
        -0.013724761,
        0.009673277,
        -0.0481782,
        0.0034036185,
        0.0028629585,
        0.0048309783,
        -0.00039036918,
        -0.01329819,
        -0.1606993,
        0.018020103,
        -0.0069765192,
        -0.009615245,
        -0.02526335,
        0.044941805,
        -0.013500418,
        -0.012667048,
        0.0074090157,
        -0.042174645,
        -0.0069312304,
        -0.012918058,
        -0.03844808,
        0.013663436,
        -0.012714183,
        0.0020715327,
        0.013330499,
        0.006451388,
        -0.0031473024,
        -0.0029549403,
        -0.022384496,
        -0.023661165,
        -0.027691372,
        -0.017557932,
        0.008159866,
        0.0016503779,
        0.030449677,
        -0.030097155,
        -0.0011776808,
        0.012944218,
        -0.0072415317,
        0.016749218,
        -0.00908007,
        0.015407964,
        0.01634559,
        0.014870139,
        -0.009180388,
        0.013293549,
        0.011849461,
        0.0071309353,
        0.0029144585,
        -0.0089311255,
        0.0034783508,
        0.0022714077,
        0.0021450042,
        -0.002932577,
        -0.012040694,
        -0.0042868312,
        -0.02980334,
        -0.0049180007,
        0.009243876,
        0.011614446,
        -0.006475046,
        0.0067189937,
        -0.010144024,
        -0.011577257,
        0.0047738403,
        0.0072209346,
        -0.02140436,
        0.0063664317,
        0.008032396,
        -0.0009480348,
        -0.12442573,
        -0.022287356,
        -0.009628999,
        -0.019930648,
        -0.00038706692,
        0.00019623428,
        0.0005694876,
        -0.006180861,
        0.033558182,
        -0.00040134435,
        0.004811323,
        -0.0016971692,
        -0.0070278174,
        -0.022002185,
        0.022104865,
        -0.026708907,
        -0.0250892,
        0.010949381,
        -0.03731375,
        0.002156423,
        0.026849475,
        0.021344539,
        0.015918965,
        0.0067364844,
        -0.019164942,
        -0.020124959,
        0.03180951,
        0.019266361,
        0.0042314245,
        -0.004818899,
        -0.0056858677,
        -0.010870334,
        0.032937456,
        0.005285995,
        -0.0007790963,
        0.0045789923,
        -0.030162746,
        -0.0060908813,
        0.011894819,
        0.038034838,
        -0.016099393,
        -0.036930982,
        0.013328835,
        0.006500403,
        -0.006731807,
        -0.0009828086,
        -0.0002753574,
        -0.014831145,
        0.04253017,
        -0.00083439256,
        0.009264544,
        0.017359167,
        0.006275114,
        0.00024361494,
        -0.019329421,
        -0.001249798,
        0.01876708,
        0.004441128,
        -0.014358269,
        0.011826892,
        0.010424215,
        0.035421323,
        -0.013188327,
        0.03809979,
        -0.00011066656,
        0.006914072,
        -0.0031504857,
        0.006291773,
        -0.01113943,
        -0.010277214,
        -0.014946901,
        0.005150064,
        -0.018624665,
        0.0072429767,
        0.018699415,
        -0.016960848,
        -0.01773684,
        0.0060196966,
        -0.0033419507,
        0.022462767,
        -0.010251675,
        -0.0013169367,
        -0.018533725,
        0.00055624323,
        0.002812092,
        0.0061965473,
        -0.015930686,
        0.0008784099,
        -0.03759296,
        -0.0032501519,
        0.00086579786,
        -0.028761497,
        -0.018231323,
        0.0080533875,
        -0.0012600101,
        0.00717613,
        0.009024979,
        0.016159432,
        0.0028689934,
        0.018550972,
        -0.003616077,
        -0.019438535,
        -0.0082142735,
        0.0039004777,
        -0.01604044,
        -0.011597579,
        0.005072733,
        -0.0022927525,
        0.009295785,
        0.001311874,
        0.0059630224,
        0.012068891,
        -0.033616643,
        -0.01541736,
        0.0026234044,
        0.017981641,
        0.0107593285,
        -0.00045548263,
        -0.031283036,
        0.00464942,
        -0.00287308,
        0.0026103323,
        0.016990505,
        0.023289427,
        0.02322276,
        -0.028680764,
        -0.014234408,
        0.022402616,
        0.010108368,
        0.01228524,
        -0.016193189,
        0.024771588,
        -0.002291386,
        -0.0049896697,
        -0.0012578067,
        -0.01501694,
        -0.027937604,
        0.022867072,
        0.008032018,
        -0.007896328,
        0.005776306,
        0.029937126,
        -0.00944554,
        -0.02667973,
        0.016749453,
        -0.014179996,
        0.00757762,
        -0.015031717,
        -0.006797985,
        0.007925135,
        -0.011208391,
        -0.0016916206,
        0.0030225685,
        -0.011050006,
        0.018521732,
        0.0074046003,
        0.014237527,
        -0.019352656,
        -0.036447477,
        -0.0142213795,
        -0.0006708521,
        -0.06575969,
        0.024477707,
        -0.019093916,
        -0.01901204,
        -0.0038399275,
        0.025769833,
        0.016906291,
        0.026829964,
        -0.021631321,
        0.015761727,
        -0.026656302,
        -0.0069863484,
        0.02856709,
        -0.020739308,
        0.018008217,
        0.018392665,
        -0.002834385,
        -0.014152479,
        0.025570432,
        -0.002540474,
        0.006301443,
        -0.0016271744,
        0.018592294,
        0.025344675,
        0.012233762,
        0.008257126,
        0.007813418,
        0.068215735,
        -0.017700791,
        7.580635e-5,
        0.008925432,
        0.031486366,
        -0.01706174,
        -0.02100586,
        -0.0060041356,
        -0.0046155476,
        -0.0037111498,
        -0.012076348,
        0.021622952,
        -0.02746968,
        0.016594427,
        0.0043138745,
        -0.015497362,
        0.002026942,
        0.022732306,
        -0.028082073,
        -0.017209686,
        0.015109489,
        0.0004721859,
        -0.012499976,
        -0.011777742,
        0.04190699,
        0.023281684,
        0.013469039,
        0.013674018,
        -0.01256249,
        0.025773732,
        -0.011881357,
        -0.015186825,
        -0.0013206592,
        0.005050834,
        -0.020919466,
        0.021312332,
        0.009853058,
        0.006720098,
        -0.018699396,
        -0.021493277,
        -0.023735236,
        -0.031815033,
        0.031405017,
        0.0050229332,
        -0.012747872,
        0.018790206,
        -0.019142648,
        -0.003358828,
        0.01468035,
        -0.0133552,
        -0.0088022705,
        -0.016717676,
        0.03626348,
        0.018934242,
        -0.0077789472,
        0.020506294,
        0.0011576456,
        -0.018717024,
        -0.019245576,
        0.0014262652,
        -0.008685345,
        0.011371256,
        -0.013593309,
        0.005603228,
        0.027167972,
        -0.0056239464,
        0.015704567,
        -0.0022569238,
        0.018337524,
        -0.004490176,
        0.020439314,
        9.338811e-5,
        -0.01149257,
        -0.026553517,
        -0.012487585,
        -0.0014783661,
        0.0026238149,
        -0.02925311,
        0.018523427,
        -0.0113531565,
        -0.011401959,
        -0.011702924,
        -0.0035205756,
        -0.01174283,
        0.00063170824,
        0.02300747,
        -0.001026172,
        0.008489874,
        0.00085706287,
        0.0016255844,
        -0.004716982,
        -0.0068622185,
        0.006889567,
        -0.0067652436,
        -0.006480185,
        -0.046015933,
        -0.00050791126,
        -0.0028999522,
        -0.0059082867,
        0.02480904,
        -0.0037176746,
        -0.0004342439,
        0.020372495,
        0.019674353,
        0.016687592,
        -0.019149777,
        -0.018696558,
        -0.019599926,
        0.0056471606,
        0.011277849,
        0.0010910993,
        0.01574992,
        0.0006813306,
        -0.031711552,
        0.013289885,
        0.0088088615,
        0.0103020705,
        -0.011887721,
        0.024244111,
        0.012343554,
        -0.0002054615,
        -0.005819506,
        -0.041874494,
        -0.008955904,
        0.001664485,
        0.028396992,
        -0.014611782,
        0.0024505218,
        -0.012476272,
        -0.028940072,
        -0.0020377056,
        -0.013595584,
        -0.016671434,
        -0.0082551055,
        -0.017968081,
        -0.008743617,
        -0.014229736,
        0.026568295,
        0.007864201,
        0.016715385,
        -0.0055983053,
        0.012007016,
        -0.0034531548,
        -0.0029075432,
        -0.012739303,
        -0.026197148,
        -0.016519079,
        -0.01907698,
        0.0041841697,
        0.0034216742,
        -0.015665915,
        0.019028466,
        -0.015490149,
        -0.0025115446,
        0.006998733,
        0.019764895,
        0.0047990982,
        -0.01579188,
        -0.003291508,
        0.0041016457,
        -0.0128195025,
        -0.024298884,
        0.028390197,
        0.014100365,
        -0.0025706657,
        0.0058321804,
        0.032452796,
        6.246578e-5,
        -0.003438462,
        0.003185754,
        0.007186782,
        -0.017617676,
        0.005818677,
        0.0016087004,
        -0.020730266,
        0.02499437,
        0.018446343,
        0.00091960677,
        0.0094762705,
        -0.026025027,
        0.0068526412,
        -0.0033128113,
        0.029702367,
        -0.0042454754,
        -0.029600838,
        -0.021604344,
        0.0049196477,
        0.047592018,
        0.01127901,
        -0.00678772,
        -0.0010876077,
        -0.038940933,
        0.008377002,
        -0.0178547,
        -0.005047879,
        -0.02082124,
        0.0028275962,
        -0.02056125,
        -0.014554616,
        0.025258942,
        0.025278548,
        -0.02373792,
        -0.011366941,
        -0.0034951079,
        -0.016550668,
        -0.0055346233,
        -0.0026155347,
        -0.030285988,
        -0.018503567,
        0.011597253,
        0.021226631,
        -0.024929032,
        -0.00133652,
        0.0023632073,
        0.0034031277,
        0.015255042,
        -0.005849544,
        -0.0015665616,
        0.015042934,
        0.0015300001,
        -0.00051416154,
        -0.005313238,
        -0.015153458,
        0.024943946,
        -0.0006627266,
        0.011496681,
        -0.009664041,
        -0.014312112,
        0.018650178,
        0.0007153758,
        -0.0018900922,
        0.012408804,
        -0.0020096942,
        0.02992841,
        0.0253843,
        0.013149079,
        0.01787032,
        0.016055455,
        0.00018187768,
        -0.032391656,
        0.009100614,
        0.013419451,
        -0.098350674,
        -0.004607563,
        -0.00082118245,
        0.0078039835,
        -0.02096224,
        -0.019659353,
        0.021808673,
        0.0025152613,
        -0.014095711,
        0.0038797653,
        -0.0050449804,
        -0.011636607,
        -0.005907268,
        -0.014486205,
        -0.0072613833,
        -0.012110433,
        0.022532234,
        0.01563354,
        -0.006107589,
        -0.02744917,
        0.029179629,
        0.009601505,
        0.015487707,
        0.013720417,
        0.019750733,
        0.01193831,
        0.03412494,
        0.02122019,
        -0.00030880663,
        0.020173244,
        -0.011276923,
        0.025243727,
        0.010545762,
        -0.017661363,
        -0.0019679286,
        0.0005595326,
        0.004214915,
        -0.00039730087,
        0.02054824,
        0.010966952,
        -0.01106307,
        0.008673873,
        -0.007430458,
        -0.027758835,
        0.00865067,
        0.0148551585,
        -0.0008879544,
        -0.0072299577,
        0.017262762,
        0.014040629,
        -0.02930901,
        -0.005131759,
        -0.0103945555,
        0.0013025055,
        -0.023948802,
        0.01928946,
        -0.0073366016,
        0.017575478,
        0.0071348357,
        0.02137545,
        0.011568343,
        -0.001639572,
        0.00422271,
        0.037560765,
        -0.031832937,
        0.037993565,
        -0.011504967,
        0.027583258,
        0.020279538,
        0.0038323468,
        -0.031912103,
        -0.015706306,
        0.006411036,
        0.010161226,
        0.009727646,
        -0.0005962308,
        -0.04249761,
        0.038784176,
        0.0023267604,
        -0.0009706513,
        -0.03948279,
        -0.00613434,
        -0.05931787,
        -0.041219573,
        0.0241341,
        -0.031395108,
        0.005160794,
        -0.015369398,
        -0.00034214847,
        0.0031560995,
        0.0062336493,
        -0.0016224323,
        -0.00018781866,
        -0.011642486,
        -0.02398691,
        -0.012917493,
        0.016864419,
        0.015023202,
        -0.010309218,
        0.001134798,
        0.011805637,
        0.011524734,
        -0.021293318,
        -0.0033998892,
        0.030254222,
        -0.008917053,
        -0.013379457,
        -0.007089947,
        -0.015173377,
        0.02425302,
        7.11609e-5,
        0.007852858,
        -0.008731575,
        -0.16218524,
        0.003798456,
        0.018603262,
        0.021104118,
        0.018796701,
        0.0076726284,
        -0.02582801,
        0.010373033,
        -0.010093338,
        -0.045859486,
        -0.00026511555,
        -0.049736533,
        -0.029913634,
        0.0024162286,
        0.019618912,
        0.13954623,
        -0.0024567924,
        0.012177922,
        -0.009826292,
        -0.006004758,
        0.005105334,
        -0.019806862,
        -0.01945691,
        -0.0036108254,
        0.010869586,
        -0.0031823155,
        0.02861621,
        0.0035022309,
        -0.010728488,
        0.02126104,
        -0.0008052292,
        0.016002685,
        0.0014255566,
        -0.012597922,
        0.03599794,
        -0.024190528,
        -0.0018181625,
        -0.018594144,
        0.0042557656,
        -0.0025745577,
        0.03069345,
        0.016211797,
        -0.020814775,
        0.00314328,
        0.010108621,
        0.02934332,
        -0.023512956,
        0.014683691,
        -0.012874141,
        -0.0099697905,
        -0.022477834,
        -0.06340601,
        0.026004383,
        -0.0019106825,
        0.02853671,
        0.014223582,
        -0.019816441,
        0.0068983207,
        0.015112419,
        0.0155879725,
        0.010511623,
        0.009461797,
        -0.008238878,
        0.006294607,
        -0.01584397,
        -0.026995972,
        0.015115075,
        0.012566446,
        0.0083999345,
        -0.002657819,
        -0.0006622525,
        0.005465301,
        -0.0025706466,
        0.022348238,
        0.012343774,
        -0.007402403,
        0.015196965,
        0.03829557,
        0.0049349447,
        0.0067931535,
        0.0018185446,
        -0.024216503,
        0.010458371,
        -0.009258528,
        0.013297967,
        -0.017098365,
        -0.0017373348,
        -0.028933357,
        -0.005770709,
        0.030384192,
        0.005339699,
        0.016306655,
        0.0007801865,
        0.019421147,
        0.008512734,
        -0.0039737057,
        0.021677786,
        -0.0016019421,
        0.003224731,
        -0.0072196177,
        0.007114169,
        -0.0067015113,
        0.014639344,
        -0.028128523,
        -0.0007355394,
        -0.02033685,
        0.008747131,
        0.015390267,
        0.012517402,
        -0.002925256
    ],
    "ai_confidence_score": 0.9999999999999999,
    "ai_extraction_metadata": {
        "extracted_at": "2026-02-15T18:30:02.732350Z",
        "ai_model": "gemini-2.0-flash-lite",
        "extraction_method": "automated",
        "content_length": 26505,
        "url": "https:\/\/www.nytimes.com\/interactive\/2025\/08\/11\/upshot\/ai-jobs.html?campaign_id=9&emc=edit_nn_20250811&instance_id=160309&nl=the-morning&regi_id=122976029&segment_id=203666&user_id=b25c5730c89e0c73f75709d8f1254337",
        "existing_metadata": {
            "author_name": null,
            "published_at": null,
            "domain_name": null,
            "site_name": null,
            "section": null,
            "publisher": null
        }
    }
}

{
    "extracted_at": "2026-02-15T18:30:02.732350Z",
    "ai_model": "gemini-2.0-flash-lite",
    "extraction_method": "automated",
    "content_length": 26505,
    "url": "https:\/\/www.nytimes.com\/interactive\/2025\/08\/11\/upshot\/ai-jobs.html?campaign_id=9&emc=edit_nn_20250811&instance_id=160309&nl=the-morning&regi_id=122976029&segment_id=203666&user_id=b25c5730c89e0c73f75709d8f1254337",
    "existing_metadata": {
        "author_name": null,
        "published_at": null,
        "domain_name": null,
        "site_name": null,
        "section": null,
        "publisher": null
    }
}