for example, i get a response from gpt. of squeeze about what public opinion or the experience of civilization looks like in relation to some phenomenon and it allows me. it may be possible to find more generalized forms. how can i perceive and analyze it from some other angle. what if not? yes, why, if so, in general, while exhaled, but do not relax neural network models they draw quite beautiful pictures there and therefore the artist tensed up very much and said that here you are studying on our pictures from our consent. this is bad. uh, all these generative models need to be banned in order to return to real art. but here, it seems to me, here, uh, the situation is similar to when the photograph appeared. it well allowed to capture reality and the artist, as a tool for fixing reality, ceased to be needed and as a result , art. i had to look for some other more creative areas in which it is possible to reflect the inner world of the artist, as if art. in general, it can be said that it has flourished thanks to this, and

>> if you use female name of the prompt, chat gpt rights more feedback. the additional words are critical in nature about 90% of the time. if you use a traditionally or conventionally mail name to get less feedback overall and a -- male name you get less feedback overall in the higher proportion is less positive in nature. kristen: in what ways could this harm real people? >> it is a great question. real people are already being harmed by bias in the workplace. managers try to give they can to unbiased feedback and real-time. we all carry biases that of the product of our own appearances. was software that carries as biases forward has the potential to do is magnify and multiply the problem beyond what one individual in drew carey forward on their own. kristen: in terms of its inclusive language, want to show you an example we prompted it to rewrite with less bias. tells how this became more equitable and less bias in the rewrite? >> it did not. after looking at the first round of performance feedback in chat gpt we asked the bot to lies in his own writin

the more accurate the results of gpt.compose an answer to a letter, but also compose a song , a poem, moreover, it turned out that he is able to write a term paper of scripts. or a cat in different programming languages, and the quality can be considered at least satisfactory, just a week after the launch , more than a million people used this chatbot, for example, students meanwhile, the gpt chat has a number of significant flaws and limitations of the neural network. for example, he can write a plausible answer at first glance, which later turns out to be meaningless. users noticed that he did not know how to compare values ​​when he was asked whether the neural network gave out specific areas more than texas or europe, but could not decide. which of them is more in the chatbot you can not upload photos, audio and video files, and it also easily inspires and is sensitive to wording. this is especially true for issues of morality and ethics. the online publication themselves claims that the intellect is posed with a dile

almost asking chat gpt is it chat gpt. >> there's also this thing bursting. >> it's another indicatorave creativity and because of our short-term memory we have bursts in our memory and with a machine they're pretty constant. >> this weekend i gave a speech at my old college. i was speaking about crew which i was on but i never talk publicly about being on the crew team. so there's nothing online about me and crew. so i put in chat gpt just to see what it would come up write a speech, 500-word speech by anderson cooper about yale crew, and what it came up was a very generic speech about crew, but it had nothing to do with me or any words i've ever used. and i realize it's because i've never spoken publicly about it, there's no -- there's no data out there that chat gpt would have used to make it sound like it was in my voice. >> absolutely. that's exactly right. it's great at regurgitating these patterns or writing shakespeare, for example. but if you're asking it to write about a niche essay written in the voice of anderson cooper it's not going to do a good job. >> we copied some te

explain how to work with it to use well, in fact, the idea arose that how can you best explain how gptrks if you do not generate e-information or a conditional book e with the help of this technology , the book was actually written in just a few days, and well, it was written so loudly, that is, it was generated with the help of charge gpt, of course, some general idea of ​​this book was also present, what are the examples, what are uh, how in principle can this technology be used and what tasks it can cope with in the best way, and indeed this book was written literally in a few days, and this once again proves how powerful this tool is, which we can use right now, oleksandr, i would like to feel the perimeter, so to speak, behind which we can use this powerful intelligence and the issue of possible errors, in particular, it is about these or other memes . we have seen, so to speak, not always correct e-e formulations from the side of the chat-bot. well, accordingly, if it will be implemented, it is possible that it will not be more correct, but tell us about some practical applicatio

whatever its limitations, chat gpt has a future in poetry. asked it to write a haiku about megan: 10 -- "barron's roundtable". jack hough: i have said too much. i don't want to bash machines anymore. jack otter ♪ i got into debt in college and, no matter how much i paid, it followed me everywhere. so i consolidated it into a low-rate personal loan from sofi. get a personal loan with no fees, low fixed rates, and borrow up to $100k. sofi. get your money right. ♪ ♪ ♪ ♪ ♪ ♪ voltaren. the joy of movement. to you, it may just be an elevator. here goes nothing. but for a young homeowner becoming their parents, it's a learning opportunity. come on in. [ chuckles ] the more, the merrier. paris, huh? bonjour! we got any out-of-towners in the elevator? tom. it is not easy. 10th floor, huh? must be a heck of a view. okay, see how everyone else is facing this way? progressive can't save you from becoming your parents, but we can save you money when you bundle home and auto with us. okay, that was terrible. okay, let's hang back. we're gonna try that agai

whatever its limitations, chat gpt has a future in poetry.out megan: 10 -- "barron's roundtable". jack hough: i have said too much. i don't want to bash machines anymore. jack otter people remember ads with a catchy song. so to help you remember that liberty mutual customizes your home insurance, here's a little number you'll never forget. ♪ customize and save. ♪ only pay for what you need. ♪ liberty. liberty. liberty. liberty. ♪ i have moderate to severe ulcerative colitis. so i'm taking zeposia, a once-daily pill. because i won't let uc stop me...from being me. zeposia can help people with uc achieve and maintain remission. and has been shown to reduce symptoms in as early as 2 weeks. zeposia is the first and only s1p receptor modulator approved for uc. don't take zeposia if you had a heart attack, chest pain, stroke or mini-stroke, heart failure in the last 6 months, irregular or abnormal heartbeat, if you have untreated sleep apnea, or take maois. zeposia may cause serious side effects including infections that can be life threatening and

whatever its limitations, chat gpt has a future in poetry.e a haiku about megan: 10 -- "barron's roundtable". jack hough: i have said too much. i don't want to bash machines anymore. jack otter meet three moms who each like to bank their own way. luckily they've all got chase. smart bankers. convenient tools. one bank with the power of both. chase. make more of what's yours. sometimes you're so busy taking care of everyone else you don't do enough for yourself, or your mouth. but eventually, it will remind you. when it does, aspen dental is here for you. we offer the custom dental treatments you need, all under one roof, right nearby. so we can bring more life to your smile... and more smile to your life... affordably. new patients without insurance can get a free complete exam and x-rays, and 20 percent off treatment plans. schedule your appointment today. jack otter: in a previous life you cold cold people with hot investment tips. you have one today? jack hough: yields tract bond yields, there is one that schwab's calls it's purchased mone

whatever its limitations, chat gpt has a future in poetry.iku about megan: 10 -- "barron's roundtable". jack hough: i have said too much. i don't want to bash machines anymore. jack otter: jack has a hot tea -- tip guaranteed to make you money, stay with us. verage. -excess cash flow. gold. the world is full of financial noise. you are right on track to hit your goals. our easy to use investing app and local advisors can help you stay on track. j.p. morgan wealth management. [coughing] hi, susan. honey. yeah. i respect that. but that cough looks pretty bad. try this robitussin honey. the real honey you love, plus the powerful cough relief you need. mind if i root through your trash? robitussin. the only brand with real honeyand elderberry. sometimes you're so busy taking care of everyone else you don't do enough for yourself, or your mouth. but eventually, it will remind you. when it does, aspen dental is here for you. we offer the custom dental treatments you need, all under one roof, right nearby. so we can bring more life to your smile...

this is a gpt and midori neural network.e choice fell on them and in general , tell us how it works at first glance, when you come to the exhibition, it seems that here the entire exhibition was made from the dirt between them, but this is just at first glance. so it seems, and gpt is a large linguistic model, into which a huge amount of data is loaded, and about 405 gb of textual information for understanding - this is about plus or minus 140,000 books and due to, uh, the information received. chad gpt can give out textual information in the form of stories and just like software code. and she generates an answer based on your request, using this linguistic model and giving you, uh, how she sees the concept of the answer to the word gpt chat 40% helped us make our press. the second network is already meters. it is based on a huge amount of information in the form of pictures to study and makes queries, selecting each element from a huge array. i don't even pretend there. how many gigabytes of information is loaded and gives

by now you heard of chat gpt or other apps. ai is coming on strong.e the internet, it's going change the way we live, and right now it's changing the way we learn. is it considered cheat field goal students use ai to answer or generate their assignments? a lot of companies investing in ai. just this week, microsoft acknowledged the ai filter isn't always working as assigned. sometimes giving wrong information and insulting users. but the company said it will continue to make improvements. the company that helped microsoft develop its own search engine is open ai based here in san francisco. it also happens to be -- chat gpt, that started three months ago, sparked widespread concerns about cheating in high school. students can type their essay prompts and chat gpt will spit out a response, essentially writing their papers for them within seconds. >> should people be worried about what this might do to the educational system? >> absolutely, yes. if we can't evaluate on a basic level whether they have learned what we taught them, then we're going to be i

this will be a breakthrough given how the gpt chatbot works.can be applied in any field of medicine education. it will be a massive trend, the largest companies will start using this high-tech platform. commit in addition, google are going to introduce the chinese canoe. and the whole russian compass i give you 2.0. well, the new neural network will become part of alice's search by mail and other basic services. the thing that she knows how to do, she knows how to analyze texts, understand the meaning , adequately answer and solve problems, the concept that we help to understand, help somehow quickly summarize, give out information as quickly as possible, launch one, launch another, then slowly, we improve all these technologies. they have already reached some, well, acceptable, as i think, level, this thing, which can bring these technologies to a new level, one of the most anticipated software development updates in recent years g54 is planned disabling microsoft products should create a new fierce competition in the field. internet search su

notice the small print at the end of the actual letter which read, "pair fry'sed from open ai's chat gpt ai language model, personal communication, februar" whoops! the school has since apologized. dumb and dumber. well, they obviously didn't read the piece recently published in the atlantic by ian bogus title "chat gpt is dumber than you think." in a moment, we'll explain how it's not just dumber than you think but more biased than you think as well. but in a statement, vanderbilt's dean of the education school, kamala benbow said her office is, conducting a complete review of the sequence of events that led to sending the e-mail and noted that the two deanlets will "step back from their responsibilities." step back? one dean said that although she stood by the bot's sentiment urging that inclusiveness -- well, she did admit to plage requiring and apologized. ok. how -- plagiarized and apologized. ok, how do they even have a job at the university. vanderbilt kicked the christians off of the campus unless nonchristians were allowed to lead the group. this is the brilliance at vanderbilt

gpt has just introduced a i tax classifier. however, in a blog post, the disclaimers that the classifier isn't always accurate and ai generated text can be edited easily to evade the classifier. regardless new technology is driving change. barbie on the academic world. ai and robots have already permeated many facets of life as we know it. >> i think chat gpt and these other ai are allowing us just different doorways and pathways to access information. >> access creativity adapting to an ai enhanced world is a new human challenge. we're in a pivotal moment right now. it feels to me actually kind of like the very early days of widespread adoption of the internet. in terms of how impactful this could be and sort of everyday life. >> that was nancy loo reporting for us tonight. coming up, camera ready for the feed. >> the top tourist spots, california have been rated the most instagram that's straight ahead. (steven) every time i come to see caremore, they go above and beyond to take care of me. i feel a lot better now. i'm taking

gpt.bard will be launched as a companywide dog food. that is silicon valley speak for an all hands on deck effort to try out the product and russia to the public in the coming weeks. big tech continues this ai dogfight, a princeton computer science student, edward has already foiled chet gpt, he was the guy who created and apps called gpt 0 to identify if a body of work was written by chat gpt. he coded over his winter holiday break and to know if something was written by human or bought. microsoft news, that student joins us live. what is your reaction to this after google announced its test mode? >> we knew this was coming. we heard this was coming and no doubt microsoft will go all out and making their ai and every thing. liz: when you were an intern at microsoft how much of our conversation in the hallways, in the cafeteria centered around ai? chat gpt which is fund up to $10 billion kind of burst onto the scene in november. >> i was fortunate to work at microsoft, adjacent to a lot of

gpt now in response to those concerns, the maker of chat gpt , san francisco based open ai has unveiledool meant to help teachers figure out if that essay was written by a student or a computer algorithm, joining us now to talk more about the new ai text classifier is victor li, a professor of education at stanford university. victor good to see you. thank you for coming on. so how will this new tool from chat gpt? help educators figure out if students are turning in work that's been plagiarized. well thanks so much for having me. i mean, there is a question as to can we tell what text may have been originally produced from the technology. but this complication in terms of, uh, was their inspiration that's provided by the technology. but then the students modify it. and is that going to be something that we consider a plagiarism if this is used as a source of inspiration or something that is going to be modified later? it will alleviate some teacher concerns initially, but i think it's gonna introduce some problems down the road what you're saying so it's sort of a gray area there. i me

quite the popularity of the neural network chat and gpt in china announced several similar solutions at once. philip trofimov philip trofimov greetings are we waiting for the invasion of large voice models ? and a souvenir, but quite a lot of neural networks are also announced. however. not everything turned out to be so simple. the first serious attempt to repeat the success of chad gpt performed by chinese scientists turned out to be not very successful just a couple of hours after the launch of the project declared as an analogue, the sensational dialogue of artificial intelligence could not withstand the load and the neural network was turned off. and this despite the fact that the language model was built and trained at the university in the foundation and died. three in china number 31 in the world. it is as of last year. that is this is a very strong public university that now makes all sorts of apologies to the gums. we are scientific institutions, we do not have enough engineers. later , scientists recognized that the level of development of the neural network of the dutch f

this will be a breakthrough given how the gpt chatbot works.ed in any field of medicine education. it will be a massive trend, the largest companies will start using this high-tech platform. i'll make a smart bot, in addition to google, they are going to introduce a chinese canoe. and yandex version the russian company is called yam, 2.0. well , the new neural network will become part of alice's search for mail and other services, the base thing that she can do. she knows how to analyze texts, understand the meaning, answer adequately and solve problems. the concept is that we help to understand, help to somehow quickly self-cut, give out information as quickly as possible. we launched one, launched another, then slowly, we improve all these technologies. they have already reached some, well, acceptable, as it is. i count the levels, this one, which can take you to the next level of these technologies, one of the most anticipated software development updates in recent times gpt, 4 its planned inclusion in microsoft products should create a new

this will be a breakthrough given how a chatbot works. gpt.ed in any field of medicine education. it will be a massive trend, the largest companies will start using this high-tech platform. apart from google are going. imagine i’ll go and this is me for you two zero new neural network will become part of alice’s search by mail and other services , the basic thing that she can do. she knows how to analyze texts, understand the meaning , adequately answer and solve problems. information launched one launched another, then slowly, we improve all these technologies. they have already reached some, well, acceptable, as i think, this thing, which can bring these technologies to a new level, one of the most anticipated software development updates in recent times, gp4 its planned shutdown of microsoft products should create a new fierce competition in area. internet search such a prediction is based on the learnability of the system and the information entered by users is simultaneously applied to product improvement is very important. uh, which of t

now gpt tea from the open ai company, it is supposed that this will make the same alice smarter, allowo better navigate in the context of the conversation, and in the classical search, the new language model assumes detailed related answers to user requests, that is, not a list of links, but a kind of source research results, on the basis of which the neural network generates some text that is readable and contains comprehensive information. well, or something close to that, saying neural networks are involved in the formation of the search results of any major modern search engine, where right on the first page, relevant excerpts from wikipedia or, for example, widgets of the necessary services, what is called an enriched response. but here neural networks play a supporting role. now we are talking about artificial intelligence. as an expert formulating an answer to the user, a question arises accordingly. how much to believe such a neuroexpert and where does he get his knowledge from, it is obvious that in order to use the system in services, we must develop a system exactly which so

this information is mostly false, i counted more than 10 errors 11:1 should check these results jet gptze where is fake news and where is real news and he will make this news on the basis of this information that it to better understand what artificial intelligence is, for example, you can ask if i can make our dialogue public. well, in principle, it is allowed to make it public, but there are certain reservations. the representative of the ministry of digital transformation, yuliya parkhomenko, says that despite the error, the chat can be used for work , some 20-page report and you are please ask, can i have such a thesis, very concise information, the most important thing, that is, it gives something like a summary of this text, it is also very convenient if you need, for example, as programmers to write some code or leave, it can also be he can also analyze the already existing program code and find some problematic ones, and some bugs in it, sometimes you have to use it carefully - says cyber security expert dmytro smobchenko password 1 2 3 4 5 reliable you don't need to write any r

talking about chat gpt. let's talk about the microsoft chat bot with reports that the messages could be threatening is that what you've heard or have read from the other reporters? >> certainly and at one point, the bot compared one journalist to hitler after the reporter kept asking questions and as you mentioned, the company helped them create a chat bot. this same company that created chat gpt. the chat bot i spent the day talking to. i asked about the ongoing controversy with the being chat bot. at one point it seemed a little offended i would insinuate the fact that you've got the same company that created you. could you potentially be capable of doing this? chat gpt told me that suspending the chat bot would actually be an appropriate response which microsoft hasn't done. interestingly, it told me that the company decides not to temporarily suspend the chat bot if it makes threats to users that might suggest the company has a lack of concern for the well-being of experienced users. >> so some competiti

so gpt, the pros and cons of using it. is the high accuracy of answers, the second is quick responses, third chatbot, tireless specialists, can work around the clock. now i began to answer you with words generated by artificial intelligence itself. it's just the same question. the producers and i asked the chat bod itself. here is the original assignment. write a text for ntv about the pros and cons of your work, and along the way, he describes the pluses they have already heard and highlights his two minuses, the lack of emotional intelligence, so he cannot always understand a person’s mood. well, the second minus the need for constant support, but in any case, gpt continues to make noise with his speech writing services. yes, on the eve of one of the deputies of the russian state duma before him , a member of the european parliament from germany, with the help of the bot, also write theses. judges decide in civil cases even this, if it is so news, in the end, where does all this lead? well, look, here are the dollar-doll

documentos legales, programas informÁticos y ensayos, esto lo puede redactar con chat gpt.ot no es confiable. >>> en la universidad de minnesota lo sometieron aprueba y pasÓ el examen de leyes pero no con altas notificaciones. ya lo han prohibido sus dispositivos si piensa utilizarlo para hacer tareas. >>> cuidado los muchachos, cuando escriban sus ensayos, originados en este tipo de herramientas, porque no es tan difÍcil determinar de dÓnde vinieron. >>> entre los trabajos que chat gpt puede reemplazar, programadores, ingenieros informÁticos asistentes legales y periodistas. ojo! aÚn no di cierne laver cid de la informaciÓn, por lo que no estamos ahÍ. >>> 5 mujeres perdieron la vida un panorama que parece empeorar si tenemos en cuenta que veamos. >>> no sÉ muy bien. raimundo Àlvarez no encuentra consuelo. 2 de sus nietas, ambas primas y una amiga de ellas cristi spi Ños fueron asesinada fernandes la comunidad de travesÍa en la zona norte de honduras. >>> yo fui la que criÉ a mis nietas, padre y madre para ellas. y ahora que yo necesito de ella me las quitan. >>> el hecho o

chad gpt, performed by chinese scientists, turned out to be not very successful just a couple of hoursject declared as an analogue of the sensational dialogue in artificial intelligence. from open ai, the servers could not withstand the load and the neural network was turned off. and this despite the fact that the language model was built and trained at the university in foundation university number three in china number 31 in the world, he finally last year. that is, it is a very strong state university, which now apologizes in every possible way. odessa we are scientific institutions. we don't have enough engineers. later , scientists recognized that the level of development of the neural network itself. dacha from the gp there was far away, and it was more of an academic laboratory experience, however, at least one moment. they have proven with success an incredible interest in technology in general reporting to the local press about creation of the first chinese analogue. cha-gper dispersed with the purpose of the network, voevo, tens of millions of posts, these users. actually, th

we were all surprised when chat gpt was writing funny songs and smart commentary on things. we will be surprised again and again. >> there a school officials right now at high schools and colleges that are trying to develop technologies or deploy technologies that can spot the fake from the real. do you think as ai develops, there are abilities to detect aiava? -- ai? >> i think the detectives are going to lose. the capability to use the technology will outpace. >> i have a devil's advocate question. can this also be used for good? let's say i really care about renewable technologies, the second amendment. couldn't this technology be used to help me get my opinion to legislators and as you describe, help me figure out the right people, the most important people, to get my opinions to? >> i think that is right. i think we want that. a tool that uses -- helps people write or translate or put tir ideas down would be phenomenal. to the extent these tools help humans, it is good for society, good for democracy. where goes wrong is where it supplants humans. where it is a million

google launched it 2 years ago, but never developed it to the state of the super-popular cha-gpt.e advertisement, our broadcast will continue with a special report by ekaterina sanders dedicated to the work russian scientists under western sanctions. diseases polyoxidonium when others do not work engine oil in continuous motion is such a hole, original japanese quality for your car in quality tape 250 g 389 rub. if you get up more than twice at night for early treatment of prostate adenoma.

and if we return to the russian student, who told in some detail how he added part of gpt to writinghis success is actually determined, among other things, by the fact that he obviously understood what and how they ask, what needs to be redone and asked again, and left him for this 23 hours of pure time, by the way, last week at the australian national university they announced that they are not afraid of neural networks in term papers and will simply not pay more attention to tasty exams. there was some increase in anxiety in the twenty-second we went to an even higher drink, because her own man appeared worried, then her grandmother died constant stress, if there is too much anxiety, if she covers, switch from empty thoughts to some real actions in this house , then he experiences tremendous physical discomfort. the study was devoted to the study of cognitive features. this causes the most anxiety and stress in people. this is the level of uncertainty our fears and our experiences. and how we treat them. today the wind blows out, they shoot a little here. well, even without this, t

to 60% free shipping. credit card alpha-banks the hardware consequences of the popularity of part of gptaccordingly, increased investments in similar solutions on the market, expect new stages of processor shortages, this time graphically, first of all, from nvidia, it is its solution that is used mostly in data centers, that is, supercomputers, in charge of training neural networks, according to an analyst at jefferson financial corporation, as of last december, nvidia products provided 86% of the total computing power of the six largest cloud platforms, and in general, out of the ten most powerful supercomputers, exactly half were powered by nvidia components well, financiers and bankers. this industry is being closely watched, among other things, because nvidia shares for a month of media noise around the dialogue of artificial intelligence gt, which he predicts almost a revolution in the market internet searches grew by 40%. and probably. will grow further. but the availability of a video accelerator, as they say, may be the other way around. decrease the fact is that to train a neur

using chat gpt to write papers.he world of higher education, technology is changing the way students approach their studies but a new trend has emerged on college campus. students are turning to ai language models, like open ai's chat gpt to write their papers. the ai goes on to write for us that in this exclusive report, it's not exclusive, it says we'll take a closer look at this growing trend, benefits and concerns and what it means for the future of education, so stay tuned as we explore the impact of chat gpt on college students and the world of academia. it's a little bit cheesy, a little bit of news speak but overall it's not bad. educators are leaning into the ai tech rather than fighting against it. jake ward is joining us now from san francisco. and jake, you are the guy. are you horrified? that was not a terrible news intro. it wasn't great, it wouldn't past our writing test. >> it's weirdly, it's kind of normal, right, it has a sort of normal vibe to it. it sort of mimics all that stuff we say. >> it doe

gpt is just introduced a i tax classifier. however, in a blog post the disclaimers that the classifier isn't always accurate and ai generated text can be edited easily to evade the classifier. regardless new technology is driving change. barbie on the academic world. ai and robots have already permeated many facets of life as we know it. >> i think chat gpt and these other ai are allowing us just different doorways and pathways to access information. >> access creativity adapting to an ai enhanced world is a new human challenge. we're in a pivotal moment right now. it feels to me actually kind of like the very early days of widespread adoption of the internet. in terms of how impactful this could be and sort of everyday life. >> that was nancy loo reporting for us tonight. next tonight on kron, 4 news at 8, a child on the peninsula back home recovering tonight after a mountain lion attack. >> the efforts to find the animal responsible and how the boy's mother scared off that big cat. plus increasing gun control in california.

large language model technologies in general and the famous chad gpt.ft has found a new application for teaching a neural network to control robots. it is more accurate to translate the wishes of a person expressed in ordinary human language into a program code understandable to machines. in general, all these tasks have already been solved in parts with a neural network. from openai. they write program code quite successfully - it even works. as commercially available the tool is called code pilot and is built into github, the largest global source code collaboration platform. in the case of any robots, including flying drones. here one more dimension of opportunity is added immediately. this is a newly created neural network code to apply to the physical world. and the main thing is to immediately correct the result, again with human speech. moreover, the range of tasks to be solved is huge; these can be relatively abstract tasks in the examples of the published work of researchers. microsoft were. in particular, the requirements for the copter. find

it is been the talk of wall street this year thanks to chat gpt.g to a huge rally in shares of c-3 ai. but a short seller report from spruce point capital management is raising new concerns about the company and the stock price saying that the price movement is not reflecting fundamental reality. they shorted the stock back in february of last year. it is sharing the latest report with us here first on "closing bell." joining us now is ben axler from spruce point capital management. welcome to the show. good to see you. >> thank you for having me. >> so you shorted it back in february of last year. i think you're up on it. it is down from that point but it was down a whole lot more before this year and the whole hype surrounding the ai. what are you doing now >> well we're revisiting the thesis if you look back at what we said a year ago about the company's aspirations being too high, about concerns with the product and of its adopt ability and the use of capital, everything came true we scored our predictions. only thing we were off on was the mag

chat gpt ai was always an integral part of our solutions for seven years. really, making sure that employees can book stuff super, super fast, expanded things extremely fast. the recent addition takes it to the next level. i can give you some examples of how much it is becoming more powerful for users and everybody. ed: i really want to know what it is like doing a transaction with open ai. it is a closed profit business. everybody wants access to the underlying algorithms. you had access for a little while. did they charge a lot. it was this the best deal ever in your career? ariel: it is a service business similar to what you get with aws and stuff like that. we are at the beginning. so, we need to see how liam would and output of the data will work. we still don't know and i think they still don't know. ed: we described what you offer at the beginning of the segment. respectfully, it is a business transaction heavy service. how has the underlying tech approved your offering? ariel: we have always had a chatbots where you ask the bot to book your entire t

chat gpt is a chat bought trade using the gpt three language model.ge processing to generate responses to user inputs in a conversational manner. not so conversational so far. let's try another question. are you alive? no, i'm a virtual assistant created by a computer programme. i do not have consciousness of the ability to feel emotions like a human. i exist only to provide information and assist with task within the limitations of my programming. zoe, that is absolutely extraordinary. some deep questions there. but in terms of you saying i should be worried, i am worried. i'm not the only one that's worried though. just talk us through the concerns with this. there are a few issues with it. number one, it is so good that there are already examples of students using it to write their homework for them. if i was a student right now i think i be delighted to know that you could literally say, write me 500 words about the end of the second world war and you will get them in about a minutes�* time, undera minute. teachers are saying well, we can spot tex

chad gpt. this neural network already knows how to write poetry posts for social networks and even program something. such a neural network, if very simply, is a black box where a person uploads billions, for example, pictures or texts, and artificial intelligence in the form of a supercomputer itself writes in them relationships. and how a child learns to write, read, draw mountains. i don’t know, for example, i don’t know the caucasus in the style of aivazovsky, he never painted, but let’s see , well, it’s interesting in a few seconds a neural network can draw this or it seems like an artist, we need to start to worry seriously neural networks are already answering calls for us in test mode enter cars on aerial photographs looking for missing people. and most importantly, they are already helping to save lives, for example, based on ct scans, it is many times faster than the doctor calculates the volume of hemorrhage into the brain, and in such cases every minute counts, the doctor spends a l

when businesses need a tool they do not need chat gpt.y need something appropriate and custom for them. caroline: they also do not need the risk of people sharing internal information that gets spit out in an inquiry down the line. there is a reporting that amazon took similar view, the general councils were that certain types of conversation looked remarkably like sensitive information of theirs. we have talked about, the cyber risk tilt into this large language models. this -- built into this large language models and generative ai. ed: the content risks as well, we reported about the inaccuracies the ai tools are generating. think about how bart's long -- bard's lunch week. -- launch week. caroline: we continue to dive into the language models i am still trying to work out what llama stands for. ed: why not llama? caroline: i love it, we are turning to cryptocurrencies and discuss the outlook for investments in agile assets. jill talks about all things the infrastructure of all -- overall crystal -- overall cryptocurrencies and the eff

the being which will include chat gpt technology.— being which will include chat gpt technology. - really about using natural language to answer the question that up until now we have been trained to ask questions in a way that we think the search engine is going to understand us. engine is going to understand us. ., .. engine is going to understand us. ., ., . engine is going to understand us. . , ., us. the room which became a household — us. the room which became a household name _ us. the room which became a household name during - us. the room which became a household name during the i household name during the pandemic is laying off 1300 staff, affecting about 15% of its workforce which has seen profits fall and its user growth slowed —— zoom. and spacex's latest issue polaris dawn will see garments like watches impacting the effect of space travel on the astronauts. fenix seven will allow people to monitor and commitment calalang —— collect various measurements. being here really reminds me of when i was younger and coming for riding lessons but i have not been on a horse

we investigate gpt and why sports in the bay area rethink how they run their classroom >>> we are learning a lot about oakland's mayor despite a lot of criticism, she is not backing down. mayor shanks out tells us the back story on why she fired the chief. >> out backlash, which is also

network, the most advanced in the world at that time and built on its improved version of the chia gptl intelligence dialog, a media success that turned out to be , well, quite unbelievable enough to remember that a site on which there is nothing but a text communication window with a robot gained 100 million active users in 2 months. this is an absolute world record for comparing those who perch for the same result took nine months and the most complex algorithm to keep users in an endless tape of b music videos. actually, all this noise and the obvious interest of the audience in order to get from robots a digestible and exhaustive answer to the formulated human language question and spurred the main players. first of all, the search market. yes, active actions such developments were obviously carried out by everyone and for quite a long time. but now he had to act quickly. if not to get ahead, then at least not very much. leave behind, including in the media sense, that is, to show that not only macros about new technologies and voice models are used in applied tasks. in general , a

chat gpt only five.means that people have a lot more access to phones, technology, maybe getting more comfortable with trying new things or looking at new things like this? >> look, i think that's also a testament to how good chat gpt is in its early days relative to the other platforms. give credit if more people want to use it, it is offering them something of value. some of the same concerns about chat gpt taken to the next level. it is a substitute for people learning how to write on their own because as you are well aware, schools are now actually seeing a problem where kids are submitting he essays on chat gpt. students need to learn to write to teach them how to think. i had an 11th grade english teacher the way you know what you think is if you can't write it down you don't know what you actually think. language carves the channels through which thought flows. when a.i. is a substitute we might get better products in the end but lose the human faculties and capacities to allow us to be thinking f

and it's a shaq gpt and artificial intelligence. it's only as good as the information going in of which someone dow posted on my business magazines. he did a search for how to make meth and distribute it in europe and came back with the correct answer. it's a very, very scary technology for college isn't for students right now. lot of rules need to be put in place fast. >> alright, hard to be the bar. but if anybody can do it might be. google will see. rob, thank you as always. on this tuesday. we'll see you again tomorrow. get your questions and comments in now. let rob know you want to talk about facebook, twitter, about facebook, twitter, e-mail directly, rob rob black ♪ ♪ ♪ ♪ ♪ you can beat it! ♪ visit youcanbeatit.org or call 833-422-4255 to ask for medication to treat covid-19. 9.25. the time for your health this morning. major changes in the works now for school lunch menus across the country. the usda putting out new guidelines for school lunches that will limit adding sugars to less than 10% of the total calories per week f

now the same part of gpt.lution from the same open ai is built into microsoft's bent search service, that is, it should work with a large number of requests per minute and not only with a two- year-old copy of the internet, but with an entire array, new information appears every day, according to experts from the financial service investing.com . a new version. the neural network has already been trained on 25,000 specialized video accelerators. let's get half the trouble bing is not the most popular search engine, but it challenged google, and of course, in its development of interactive artificial intelligence, it urgently accelerated. and let the presentation of the project called bard. it turned out to be a hasty seeker. sooner or later , this neural chat will be launched without fail, so it was calculated on the market and it turned out that if, at the current level of effectiveness of such neuro dialogue bots, each google search query will be processed by an analogue of cha-g5, then the corporation will

i created a plan, uh, which uh, later ran through the chat gpt.les of management, and gives me. uh, the answer is five to thirty managements detailed. then i see that something needs to be corrected here. i correct it myself or ask, but to provide a different answer instead of thousands of requests on the search site alexander asked her roseti only a few dozen, and then brought all the answers to a common denominator no study of volumes in the scientific literature , analysis of the search for patterns and conclusions on writing, taking into account the primary editing 15 hours, and for editing after the comments of the supervisor for another nine hours, that is, in total , the attestation commission did not notice a dirty trick for about a day. the diploma was tested for uniqueness with an indicator of 82% alexander, although it was a three, but defended it, and the gpt chat artificial intelligence programs have already managed to make a lot of noise around the world, the development of the american company open ai, appeared at the end of last ye

intelligence, everyone is enthusiastically writing at the event about such and such a program chat gptw if i am correct in rejecting this abbreviation, but for the audience i will explain that it is such a program that helps to write texts and can communicate and imitate actually a person , a very educated person who is erudite and knows a lot of interesting things, the only difference is that this is not actually a person, but a program, and even some ukrainians use it to collect information, for example , to analyze what is written about ukraine in western mass media, can we say that we have already reached such a level of development that artificial intelligence somewhere can even deprive someone of his job? well, mr. mykhailo, we talked with you , let us give your comment on this , to predict what the future will be, especially for a long period of time. difficult if at all likely but there is such a simple rule eh and it is logical that we tend to exaggerate the impact of technologies in the short term and underestimate their impact on our lives in a longer period eh what does thi

the gpt chat happens in the format. chat itself and its capabilities amaze users around the world. only can he consciously communicate almost like a human, but he generates huge and unique artistic and technical texts in seconds. the fight of complexity in any style is able to program , translate and try on roles and characters. he successfully passes his exams manager's lawyer. so the first thing alexander asked for was electronic gene write me a diploma gpse chat, and on request write a diploma, but refuses to fulfill it and this is definitely good, and therefore, in order to write the same diploma, i, but thought out the algorithm that i i will see in this diploma and the neural network offers me options. for example, i ask you to write five principles of modern management, and then i ask each of these principles to expand, and examples from the literature. also i can ask replace stats from english sources with russian sources intelligence, alexandra and artificial intelligence compiled a diploma for a total of 24 hours, the neural network wrote, alexander edited drew lotal tabl