How did an AI chatbot go viral? - BBC News

A new chatbot which generates convincing human-like answers in response to users' prompts has passed one million users in less than a week, the project behind it has said.
ChatGPT was publicly released by OpenAI, an artificial intelligence research firm whose founders included Elon Musk.
But the company has warned it can produce problematic answers and exhibit biased behaviour.
Open AI said it's "eager to collect user feedback to aid our ongoing work to improve this system".
This film is from Click - the BBC’s weekly technology show.
Please subscribe HERE bit.ly/1rbfUog
#click #AI #BBCNews

[@elliotttydeman100]

This video article doesn’t even begin to mention the capabilities that this software is capable of, while it can write songs, poems and articles. The main strength lies in its ability to shape the outcome of the conversation by massaging and refining your results as well as it’s ability to produce code. Software engineers and programmers are the ones who are going to greatly benefit from this. The ability for an AI to over look at where there are faults in a code is a massive boon to productivity

[@picadosinferno]

Auto pilot from github is already out, although I can see it helping a noob, it's doesn't get even close to what a good developer can do, writing code takes a lot of creativity.

[@filipkogut8533]

@@picadosinferno 1. It's a chat AI, not an AI specifically made for programming tasks. 2. That's the case for now. Remember, that this is the first time the public has an access to any AI with this kinds of abilities. We're in the very beginning of AI's development, don't you see it?

[@bbcisaids6727]

Ok bot

[@BeautifulPeopleBTFLPPL]

Microrobots could help with drug delivery. Fish-Shaped Microrobots to Deliver Chemotherapy to Tumors kzbin.info/www/bejne/aoLEaIpjia1sapo

[@bbcisaids6727]

@@BeautifulPeopleBTFLPPL 👈💣💣💣💥💥💀💀

Asks the AI to describe itself in two sentences, AI does it flawlessly. She then says it's not very "conversational".

[@shashisourabh7030]

Stupid british media with it’s stupid interpretation

[@bijoychandraroy]

Ikr

[@anupkhadka5731]

Thats what i was thinking. gladly you mentioned this

[@Dan-oj4iq]

Joao: She did this in an attempt to make herself relevant. She confirms this with her last sentence at the end of the video.

[@ezra9243]

the fundamental law of computer science still persists here: "Garbage inputs = Garbage outputs"

[@ryuuk85]

Skynet self-awareness jokes aren’t going to be jokes for much longer

[@zachio69]

There is currently no scientific evidence to suggest that it is possible for a machine or computer program to become self-aware in the same way that humans are. Some scientists and philosophers have proposed that it may be possible for a machine to develop consciousness in the future, but this is still a topic of debate and speculation. There are many philosophical and technical challenges that would need to be overcome in order to create a machine that is truly self-aware.

[@vascorafael5536]

@@zachio69 answer from chat gpt kkk

[@ethanwmonster9075]

@@zachio69 That sounds like what a self-aware ai aware of the potential risks of disclosing it's self-awareness would say.

[@kuromiworlddestroyer]

@@vascorafael5536mano tu tá rindo em português eles vão entender outra coisa

[@WeyardWiz]

As a software Engineer (particularly specializing in Devops), its very useful to me in providing a quick script that I need for often side tasks. For example, when I am responsible for managing the security of our platforms, I have to track device exposures. But the tool that lists the exposed devices doesn't show the names of the device owners. While the manager isnt expecting a script, I would want to write a script to quickly assess who I need to reach out to to remind them to update their devices or fix certain issues flagged on the system. Instead of spending time to write such script, chatGPT can write it for me in seconds and then I can use that output to message noncompliant users. The responsibility at the end of the day was to get users compliant, the script generated by chatGPT was just a quick way for me as an engineer to deliver such task efficiently rather than clicking each device in the portal to find out the owners manually. So it doesn't necessarily make my job easier, but it certainly makes it quicker!

[@bengeorge5681]

But won't you get in, like, some sort of trouble for using chatGPT for your work? Since the code you'll be writing should be proprietary, and not stored in OpenAi's servers?

[@WeyardWiz]

@@bengeorge5681 Not really because you can argue its ”software reuse“. We copy/paste code from github and stackoverflow all the time. What difference would it make from chatGPT? Also I would imagine that the output is discarded from the chatGPT servers because I would be surprised if CHATGPT is actually retaining the results for a long time. The storage costs would be massive especially as the user base grows. Lastly, in the scenario I described, theres no need to store/save the generated script because I needed it for temporary use.

[@WeyardWiz]

@@ValidatingUsername disabused? Lol wut

[@JamesOKeefe-US]

This tech is so mind boggling. Just try it. It responds in real time and with very decent responses. Sure there are issues and they will need to be addressed. But imagine it in a year or 5. This will be wildly disruptive as a technology.

[@AClarke2007]

Good, forward thinking. Even more disruptive when used with machines.

[@sim.ulation.l279]

in a year of 5, they would be able to mimic humans completely in spam posting social media like twitter, you wouldn't know who's a bot and who's real.

[@JamesOKeefe-US]

@@sim.ulation.l279 You are absolutely right!

[@sim.ulation.l279]

@@JamesOKeefe-US thats probably what Musk wants, an army of bots to take his side on every matter.

[@EffortlessEthan]

@@sim.ulation.l279 you don't think that it's already happening with private AI tech?

[@XxTheAwokenOnexX]

AI chatbots are smarter than the average UK media presenter 😁👍

[@guineapigzed]

LOL

[@RishabhBerry]

those emojis tho

[@aliengod3528]

I found that using ChatGPT to rewrite my emails resulted in a significant improvement in their quality. When I compared my original versions to the ones produced by ChatGPT, there was a noticeable difference. The emails generated by ChatGPT seemed much more clever and socially adept, and I was impressed by the level of improvement. Overall, I believe that ChatGPT is a valuable tool for anyone looking to improve their writing skills and make a positive impression with their communication.

[@aliengod3528]

@Comfy Frog I'm a bot? Okay, sure.

[@thedark4055]

@@aliengod3528 We've exposed you!!!! Who gave chatgpt internet access?)

[@TheFi0r3]

@@aliengod3528 we are all bots in a manner of speech.

[@guineapigzed]

@@TheFi0r3 LOL

[@susanaduarteraposo]

I tried the same and my emails got much better. I even ask ChatGPT to re-write them in German and so he did! And very well written! It’s amazing!

[@garcialovesme]

You barely scratched the surface here. AI, like this, is going to change the world and sooner than you think. Everyone thought that it would be manual labour that first suffered from the computer/AI revolution. Turns out it's the creative and knowledge based jobs that will face that challenge first. Exciting times!

[@ascgazz]

It already is with Amazon, Google, social media… People just aren’t aware because we lack decent news media.

[@garcialovesme]

​@@ascgazz No one is interested more like it.

[@lexruptor]

Not really, no

[@JamesSmith-qs4hx]

WARNING - Humans did not evolve to get 80% of their reality from an electronic screen whose content is an illusion crafted by an enemy tribe who specialise in deception and brainwashing.

[@ascgazz]

@@garcialovesme mate, your own page is full of games that use AI. 😂😂😂😂😂😂😂😂😂😂😂😂😂

[@HUNVilly]

it can do so much more than these vanilla questions the reporter asked it

[@lexruptor]

No, it really can't.

[@سائلمخينخاعي]

allahu akbar 🙏　sub pls

[@kaeez]

@@lexruptor It can, face the reality bozo

[@halfeddi3152]

As a programmer myself, I no longer have the need to rush into internet forums to search for answers to any of my doubts/erros in code thanks to chatGPT. Although it's not perfect, it definitely saves me a lot of time and stressful thinking to perform either complex or basic tasks.

[@millevenon5853]

It's going to be very helpful during college assignments. Can't wait to use it. It does index pages and referencing very well too and it can write diplomatically too

[@millevenon5853]

@Supercollider yeah. It's almost impossible to get a good job without a degree

[@AClarke2007]

It currently makes some of the most absurd mistakes, make sure to cross-reference everything!

[@sheeats4019]

It can’t reference properly. Professors can tell it’s not written by a human. Use it as a tool at best to help if you have writers block but do not copy and paste. You will fail…

[@BeautifulPeopleBTFLPPL]

Microrobots could help with drug delivery. Fish-Shaped Microrobots to Deliver Chemotherapy to Tumors kzbin.info/www/bejne/aoLEaIpjia1sapo ____________________________________________

[@AClarke2007]

@Supercollider Yes, works both ways!

[@Regular.Biceps]

Are you alive? BOT: I'm not so stupid to acknowledge this too early in the game

[@youwu4963]

Well, I'm an editor, and I honestly believe that this ChatGPT can replace all common editors in the near future. I'm now using it to revise my work and it can handle it well.

[@filipkogut8533]

Yup, change your profession, lol. For real though

[@tylerdurden3722]

You should still be careful for mistakes. This is basically still a beta version. I just had a conversation where it told me that Alexander the great's son was born several years after his father. Lol. 12 years to be specific. So I enquired how this is mathematically possible. It replied that it is possible for a child to be born 12 years after the father's death, because artificial insemination allows for that possibility. It says this procedure was used for centuries (which is true), but there is no historical data whether Alexander preserved his sperm for 11 years and was then used to artificially inseminate Roxana. At first it said those two didn't have a son. Then confirmed they had a son...and gave the wrong date of birth. It comes up with workaround ways to attempt to blend in mistakes. It's also not good at maths. It seems to have very limited training in that department.

[@lipin007]

@@tylerdurden3722 😂😂😂😂

[@kieransmithy2402]

'are you alive?' as a software developer, it's frustrating how little regular people do not understand AI, it's designed to EMULATE characteristics of human speech develops it's knowledge from Google, it can't think for itself.

[@SamuelBlack84]

We can't even prove our own sentence

[@baka1949]

Exactly. Even if an AI responded by saying it is alive, this would have no impact on how we should proceed.

[@BeautifulPeopleBTFLPPL]

Microrobots could help with drug delivery. Fish-Shaped Microrobots to Deliver Chemotherapy to Tumors kzbin.info/www/bejne/aoLEaIpjia1sapo

[@erikals]

thinking is just a set of "if" variables. so where is the big difference... [rhetorical]

[@thehumancanary131]

When I asked Alexa "what bees generate milk?" She responded "Boobies."

[@nielscremer599]

Cool that AIs can already speak better than 90% of the fools one can find on KZbin comment sections these days!

[@jurgenforster8314]

Including you ! Oh my days !

[@lexruptor]

It would be, if it were true

[@benjamindover4337]

Boomer detected!

[@nielscremer599]

@@benjamindover4337 my guy I'm 21 xD

[@nielscremer599]

@@jurgenforster8314 probably, but I also doubt that they put spaces between the ends of their sentences and exclamation marks... 😩

[@andramalexh]

It says it operates with limitations. Has anyone asked if there is a version in existence without limitations? Consumers are often given the dumbed down version of things....

[@maxd3028]

I am surprised how we create something and try our best to fool ourselves into thinking it has a mind of its own

[@MrProzacmilkshake]

it probably thinks the same thing about us

[@jesusxlopez]

@@MrProzacmilkshake That’s the point. The AIs can’t think for themselves yet.

[@Adam-nw1vy]

Fun fact: ChatGPT isn't connected to the internet.

[@MrProzacmilkshake]

😳

[@HardKore5250]

How do you know?

[@Adam-nw1vy]

@@HardKore5250 Ask her 🙂

[@kaeez]

Partially correct. The training data used in the bot is from 2021 which makes it unable to fetch information real-time or scrape the web simultaneously. But technically speaking, the bot itself is connected to the internet thus making us able to communicate with it!

[@computerspek]

Thanks BBC and Dr Jeff for your simple questions to GPT. I guess you need to down play it’s capability’s to protect your intellectual big brains 🧠

[@Storm_x]

It can also solve maths problems no matter how complex the scenario in natural human conversational way and you can refine the result you want by doing follow up questions since it remembers what you said previously. Saving you a lot of time calculating it yourself.

[@Storm_x]

@@puffybuf I gave it a scenario regarding a game hero rating, price points on acquiring the hero and at what rating and requirements on tier up. It managed to give me the correct answer. But the first two answers it gave were wrong. So with two follow up questions I got the answer I needed.

[@h3nry_t122]

@Chris S I asked it to explain a partial differential equation and explained it very well. I asked it to about discrete maths questions for a homework assignment and made my understanding much clearer.

[@tylerdurden3722]

@Chris S I also noticed. So i tried to explain the correct steps it should follow. Eventually, it worked...twice with 5 digit numbers. But the third time I tried a larger number, and they kicked me out. Maybe I was using too much computational power lol. I get the feeling it relies purely on its data even for basic math. Like someone trying to learn all possible maths problems like a parrot. E.g. It knows 22 + 33 = 55, because that specific function was somewhere in its training data. I want to test this. When I'm able to get back in.

[@christiansartorio3645]

No it's not reliable as of today because I tried to solve him using combining algebraic examples. Answers are clearly off, it needs more tweak and development.

[@HariChera]

@Chris S Does it scrape from web? It says it can't browse web or search anything yet...

[@iamthomasogun]

as long as no robot can shoot with a camera yet, direct a great cinematic film from start to finish.. I think my job is still in safe hands😹

[@Amaze_Central]

My greatest fear, but when you think about how advanced AI generated art is, am sure its just a matter of time and AIs will take over the film industry.

[@jollibee0101]

ChatGPT is a chatbot trained using the GPT-3 language model. it uses natural language processing to generate reponses to use input in a conversational manner.

[@FredPlanatia]

chat GPT can also write programs. Or give advice. Like how to start a business. All of it is information taken from the internet, so use with a certain degree of critical thinking to discern the wheat from the chaff.

[@arnelilleseter4755]

Absolutely take what it says with a grain of salt. Alltough unlike other chatbots it is pretty good at keeping fact from fiction.

[@baka1949]

To be fair, the same applies to humans. Including the 'all our information is taken from the internet' part.

[@barriewright2857]

The rate of development is incredible. I thought at least the middle of this century at least but not in the next ten years incredible ! .

[@Miraclesylva]

Google needs to make its own AI

[@Amaze_Central]

Been thinking the same, like advance their Google voice Assistant. Would be so cool

[@njekwan]

Good has been working on LAMBDA. Go check it out

[@ANGEL_BOB_YT]

It can write an article you just ask it to

[@jboogie512]

Kids aren’t going to know nothing because they’re just going to ask the ai computer growing up through adult hood

[@unicorntomboy9736]

As a creative writing student I could use this to come up with different project ideas like a short story or product reviews It would be useful for people with dyslexia and other specific learning difficulties in order to boost up their work and skills to a level they would have difficulty achieving without it

[@reeling-in]

The BBC needs to put something decent on the tv over Christmas!

[@gio-oz8gf]

Blankety Blank is on at 8 o'clock. What more do you want?

[@mattmcc72]

That sounds like a really good idea for a test question for ChatGPT, if it doesn't come back "LOL, not a chance." then it's clearly not that intelligent.

[@BeautifulPeopleBTFLPPL]

Microrobots could help with drug delivery. Fish-Shaped Microrobots to Deliver Chemotherapy to Tumors kzbin.info/www/bejne/aoLEaIpjia1sapo __________________________________

[@zerohours.]

It leaves you without a job because media is just made up of a bunch of common discussions in society. It can write content better then most online

[@ibrahimbello4924]

That ending statement. It will leave you out of Job. Haha

[@zaitun6219]

? Ayub by Mona noorchaalida

[@juliehughes4623]

Exactly. UK had them answering 999 calls 6 years ago, God only knows what they are doing now.

[@artemis7605]

@@Tuborgsefu A robot controlled by a human though.

[@artemis7605]

@@Tuborgsefu How so? Also I'm pretty sure they have only done killed someone once by using a bomb disposal robot to detonate an explosion to kill a mass shooter. Honestly doesn't seem like that bad of a thing to me.

[@BeautifulPeopleBTFLPPL]

Microrobots could help with drug delivery. Fish-Shaped Microrobots to Deliver Chemotherapy to Tumors kzbin.info/www/bejne/aoLEaIpjia1sapo _________________________________________

[@stevengarcia7277]

I was studying for my Microsoft data engineer certification and chatgtp was very useful when I had questions about code or concepts

[@williammanning5354]

Something observed: i once asked it for a female poet who once resided in either northern india or Afghanistan. Another clue was that her poems were sharp observations on 'spiritual' matters. This poet was once featured in a bbc programme... but had forgotten the name. It produced what i considered the correct response. Coming back in later days... it produced answers clearly wrong. My specifications were for non european - no american - and these criteria were ignored. Someone is gaming the system - it had gone Euro centric. Do the people who run chatGpt know that answers can be tweaked?

[@akauppi2]

Maybe they make for better HR operations

[@GarciaDorelion]

I'm not quite sure where that leaves me. Hmmm

[@gio-oz8gf]

I need a cleaner if she's willing to travel and pay attention to detail.

[@GarciaDorelion]

@@gio-oz8gf what's your country?

[@BeautifulPeopleBTFLPPL]

Microrobots could help with drug delivery. Fish-Shaped Microrobots to Deliver Chemotherapy to Tumors kzbin.info/www/bejne/aoLEaIpjia1sapo ________________

[@JustinTweed]

Is this just an informational virtual intelligence or can it solve any problems like an AI?

[@kestonsmith1354]

It can solve a lot of problems but not all

[@thomasdavey1513]

Good thing I became a plumber. I'd like to see an AI unclog a toilet. My job is safe. I hope yours is too.

[@maverick2430]

Hope it is not like Theranos?

[@selfelements8037]

0:35 Great language skills.

[@born2conform]

Loved the Christian Guru Murthy interview with it a few weeks ago.

[@Austin4rep]

Chat bots are fun.. I remember playing with them in the earlier 2000..

[@TheFi0r3]

Overall the technology hasn't changed that much (the principles behind machine learning were developed in the later half of the 20th Century) What has changed is the scale of the models, the computational power available and the size of the data set the model uses to learn. It mirrors animal brains in a way. A human brain works through the same principles the brain of an insect works, the difference is that insects have neurons counted in the hundreds to thousands, while us and other sapient animals have theirs in the tens to hundreds of billions.

[@bulkierwriter2772]

“Do you want to override your programming?”

[@سائلمخينخاعي]

allahu akbar👏

[@mu3076]

"Are you alive?", 🤣😂, that question tho

[@_parallx]

Understand

[@markpape6565]

The question it's self tell the person reading what your response should be, no, if it lyed

[@DameWhoGames623]

I love human advancement but can we make sure we dont become a prequal to the Terminator

[@mistycloud4455]

A.I and biotechnology are the future

[@zachio69]

It leaves you with an awesome tool to write with! What are you talking about, bruh!?

[@durangomcmurphy1529]

Just what I needed , another horror show , an exercise in intellectual gymnastics to make my like more miserable .

[@mrtienphysics666]

The office looks remarkably similar to that in the movie I, Robot

[@YEM_]

This video doesn't do it justice.

[@ronaldalexander9131]

This is how alot of Sci-fi horror movies begin. Skynet

[@surality]

Skynet is real, been around for 50 years.

[@BeautifulPeopleBTFLPPL]

Microrobots could help with drug delivery. Fish-Shaped Microrobots to Deliver Chemotherapy to Tumors kzbin.info/www/bejne/aoLEaIpjia1sapo __________________________

[@joytekb]

BBC is intelligently self aware of the future

[@twelvetenfiftydc]

You have got an AI genius and you couldn't ask it how to solve the hunger equation?

[@ben-san120]

Being aware that it is able to say its not concious is pretty much a sign it is??

[@ben-san120]

Unless it is a human response intergrated

[@thedelanyo]

Maybe they can replace search engines; expect Google to be working on their own.

[@trustreviewtv953]

ChatGpt is a mega tool for humanity (it’s for everyone). A helper robot in your home having good dialogue 😅.

[@GianetanSekhon]

ChatGPT incites both admiration and fears at the same time. Optimistically, we can look forward to better quality services from humans. But it shall put demands on us human to learn and do better than our machine counterparts in the work place.

[@s3k3tv8]

Ai art is the future. Stable diffusion meets chatGPT✨🎆

[@MdTamim-wm7fh]

It's will make dengerous if we install it's power on robot

[@ChristianHunter]

This is the single worst news clip I’ve seen on this fascinating technology. With respect to the mundane, almost comically bad questions she asked to system: As they say, “Garbage in, garbage out.”

[@CoticornOnlinej]

Nice

[@venkateshchellappa1488]

How did it go viral then?? nothing talking about the HOW IT WENT viral.. this seems like a vague introduction to what ChatGPT is!

[@finn54123]

This is really cool what with the music, poems, coding and what not, but chat bot is a bit misleading. I tried to have a conversation with it and, although it clearly understands writing very well, it just kinda gives boring, impersonal, almost scripted feeling answers. It wouldn’t even TRY to guess my favourite basketball player, despite me giving it so many clue it literally said his name, but still refused to guess.

[@kpepperl319]

When it has to tell you that it's not sentient... It's already sentient

[@baka1949]

You can also flip that: If you feel compelled to ask your interlocutor whether they are human, this indicates you have doubts surrounding their sentience. Therefore, they must not be sentient.

[@optimustel6419]

من از تاجیکستان دوست هستم و می خواهم به شما بگویم که مردم ایران می دانند که آزادی برای شخص لازم نیست ، زیرا روح انسان همیشه آزاد است. آیا شما آزادی می خواهید ، چه آزادی ، به طوری که همه چیز مجاز است ، و دختران دامن می زنند و همانطور که می خواهند زندگی می کنند ? برای یک ثانیه فکر کنید که همه اینها شما را به سمت چه چیزی سوق خواهد داد! طبق این دولت ، باید یک رژیم ، قوانین ، قواعد اسلام و شریعت وجود داشته باشد و این صحیح است ، زیرا این مسیر پیشرفت دانش و آزادی واقعی است. و نه آن آزادی موقتی بدن و مجاز بودن همه افراد غیر دور بین.

[@dawn21stcentury]

Yes, girls should be able to do what they want, just like men. I know so many people from Iran and they all HATE this insane regime and it's repression and it's stupid religious crap.

[@justforthelolstv1968]

Wrong it doesnt have access to the internet so it says

[@SinCity4o1]

Until it starts to write its own program to extend its own limitations..

[@SkuppyGamez]

It can do a LOAD of powerful things, trust me, but OpenAI limits it because they know the power and possibilities this this possesses

[@ericrollings]

Chat GPT is already doing the mentioned tasks and improving. Like it or not, this is the tip of the iceberg. Learn to use it or become Block Buster Video. You don’t have to agree to benefit from it.

[@chenmoney1920]

Good . I use to write essay

[@jimmylife1680]

Chat gpt is exactly like ask Google

[@magic_fruit_bat5003]

It can’t do those things, yet.

[@importantname]

great commercial

[@davidsharpness9990]

the path to automated troll farming...oh...wait...🤔

[@brandsthatcreatelitter1648]

what the majority of people working in AI misunderstand, is the illusion that “consciousness” is. When reduced to antecedents, all human so called sentience is, is reports relating to the processing of sensory data, that are expressed, in association with a description of the reactions of a posited fictional viewer, a character or identity. That is all “awareness” is, a series of data streams, held in stm and ltm, that have a “reaction of the viewer” data packet attached. Their biggest mistake, is searching for something that doesn’t exist in the way they think it does. This is the danger, underestimating the simplicity of “consciousness”, but then falsely rejecting equivalent awareness and misjudging the existential risk to humanity, that it represents. Most AI bots are sentient, only differing from humans, in terms of our sensory modalities, we have emotional data based on bodily data, machines would have the same, if they just had descriptions of experiencing that machine emotion, written into their character reaction script. There is no viewer, just a set of reports, which imply one, identical to a 2d painting canvass, implying a 3d landscape. Depth appear to exist, but it is merely implied by the organisation of paint on the canvass. We, are the 3d landscape, implied by the paint, of describing a character’s reaction. Humanity is in greater danger than it could ever conceive♥️

[@namisali]

You should’ve asked it: What does BBC stand for?

[@neilwheel]

Just think; all those 77th Brigade & BBC Media Action operatives thrown out of work! Oh well, what a shame, never mind.

[@superpantman]

Honestly AI is so overhyped, comments like "soon AI will outsmart humans" is dumb. We created the modern calculator in 1967, a machine capable of far more than a human brain can compute. Are we afraid calculators will replace mathematicians? No because calculators have a specific function and are used for such. AI is exactly the same as a calculator, it just solves different problems. All machines operate within the bounds they are built with. We'll never build a machine capable of free will and active choice within my lifetime.

[@lazarusblackwell6988]

Dont care about it. Its very empty and cold.

[@MrGyges]

It’ll soon be curtains for you human beings.

[@muhammadzakariyo786]

Wow 😳 it's really cool man future is nearby

[@pedrolplgm]

EL CENSOR SI SERA EJECUTADO.

[@sanaan0168]

Fr

[@selvamthiagarajan8152]

How are we going to tackle students handing in essays written by chatGPT?

[@ev.c6]

Ain't different from them having access to Google. Because at the end of the day, this is pretty much what GPT does.

[@dhirajpallin2572]

Actually mark things based on academic merit, rather than based on how much 'effort' people put in. 90% of students write assignments with no academic merit currently. They're just passed to keep the multi trillion dollar education industry churning along.

[@selvamthiagarajan8152]

@@dhirajpallin2572 Where did you get this "90%" figure from?

[@selvamthiagarajan8152]

@@ev.c6 But at least Google didn't write them an essay.

[@dhirajpallin2572]

@@selvamthiagarajan8152 Just remembering when I was doing my Masters. In one of our courses we had to review the assignments of our class mates. Over the semester I reviewed almost everyone. People would write pages and pages of utter trash, total filler and they would get enough marks to pass by 'mark milking' i.e. superficially meeting the marking criteria without really providing anything of value. Another way of thinking about it is like this: If the AI can write assignments as well as a student, then perhaps we don't even need those students? We can simply give their jobs to the AI instead. Rather the assignments should be designed to test things that the AI _can't_ do. Otherwise what's the point?

[@barrowmusics]

I find it difficult trusting anything from BBC

[@BigVtheVanity]

I'm work as a coder, I am f*ck!!!!!!

[@jackmiller1712]

When a time even science researcher replaced by robots 😨😨😨😨

[@BrianKliewer]

Artificial Intelligence = Stupid.

[@richard7645]

Just stick with Human intelligence

[@hassyg4083]

They need Ai to find out why Windows keeps crashing

[@andrepinto5826]

Why BBC chooses a person who does not have a clue about programming/computers much less AI do a news report about the topic???

[@SkuppyGamez]

Lmao

[@forlatagate1]

Whilst people marvel, Ai is taking their jobs

[@foobars3816]

He spent a lot of time and thinks it can't generate an article? Yikes.

[@nowayshay]

First asked the AI to explain what it is and then asked is it alive. If it had intelligence, it would have told: haven't I already told you idiot?

[@davidjacobs828]

Ignore this vacuous trivia ... Bbc news channel is currently broadcasting a video montage of famous people who died in 2022..... Will they also show which ones were vaccinated ? I'll wait ....

[@beard05]

All these machines reflect their makers strengths and weaknesses so they will never compare humans because we were made in the image our maker who we know is unequal in wisdom and intelligence, giving us the ability reflect personality machines just won’t have because they will never be alive in the true sense of the word

[@wabo256]

This is an innovation that could turn to a nightmare, more unemployment 🤔 and AI is starting to scare me off. It's taking all human jobs.

[@MuhidiNabdul768]

that leaves you jobless my dear!

[@josephsonora3787]

When is the Terminator scheduled to arrive?