OpenAI is developing software to detect text generated by ChatGPT
- Reference: 1673263807
- News link: https://www.theregister.co.uk/2023/01/09/in_brief_ai/
- Source link:
Reports of students using AI to do their homework for them have prompted teachers to think about how they affect education. Some have raised concerns about how language models can plagiarize existing work or allow students to cheat. Now OpenAI is reportedly working to develop "mitigations" that will help people detect text automatically generated by ChatGPT.
"We made ChatGPT available as a research preview to learn from real-world use, which we believe is a critical part of developing and deploying capable, safe AI systems. We are constantly incorporating feedback and lessons learned," a company spokesperson [1]told TechCrunch.
[2]
"We've always called for transparency around the use of AI-generated text. Our policies require that users be up-front with their audience when using our API and creative tools… We look forward to working with educators on useful solutions, and other ways to help teachers and students benefit from AI."
[3]
[4]
Being able to distinguish writing produced by a human or machine will change the way they can be used in academia. Schools would be able to enforce banning AI-generated essays more effectively, or maybe they might be more willing to accept papers if they can see how these tools can help their students.
Yes, generative language models can be good but they don't know what they're talking about
As impressive as AI-generated writing may seem in the headlines with academic conferences and schools banning machine-written papers, here's a reminder that they lack understanding compared to real human writing.
When tools like GPT-3 or ChatGPT surprise us by coming up with shockingly good responses, experts say it's proof the model is capable of encoding knowledge, but when they fail to get things right they're said to be hallucinating. Don't be fooled, Gary Smith, a professor of economics at Pomona College, reminds us.
In an op-ed [5]published in Salon, he showcased a few examples where GPT-3 fails to reason and answer questions effectively.
[6]
"If you play around with GPT-3 (and I encourage you to do so) your initial response is likely to be astonishment… You seem to be having a real conversation with a very intelligent person. However, probing deeper, you will soon discover that while GPT-3 can string words together in convincing ways, it has no idea what the words mean," he wrote.
[7]Alphabet reshuffles to meet ChatGPT threat
[8]OpenAI predicts biz can break a billion in revs by 2024
[9]Apple taps brake on self-driving cars, now aims for 2026
[10]OpenAI opens doors to ChatGPT, another AI to fill the world with kinda-true stuff
"Predicting that the word down is likely to follow the word fell does not require any understanding of what either word means – only a statistical calculation that these words often go together. Consequently, GPT-3 is prone to making authoritative statements that are utterly and completely false."
OpenAI released ChatGPT, a newer model last November that is designed to be an improvement on GPT-3, but it still suffers these same issues nonetheless – like all existing language models.
Apple is publishing audiobooks narrated by AI bots
Apple is seeking to partner with indie writers and publishers to help them narrate their books using voices synthesized by AI.
Authors were advised to reach out to Draft2Digital and Ingram CoreSource, two companies that produce and publish e-books on the Apple Books app, if they want to turn their work into audiobooks. They are only accepting submissions written in English for romance and fiction, other genres are not yet supported.
"More and more book lovers are listening to audiobooks, yet only a fraction of books are converted to audio – leaving millions of titles unheard," Apple [11]said in a blog post.
"Many authors – especially independent authors and those associated with small publishers – aren't able to create audiobooks due to the cost and complexity of production. Apple Books digital narration makes the creation of audiobooks more accessible to all, helping you meet the growing demand by making more books available for listeners to enjoy."
[12]
Compared to the robotic, tinny sounds computers used to make when they mimicked humans, synthetic AI voices have vastly improved. They now sound pretty natural and are less monotone.
The new feature will allow self-published writers to expand their audiences and gives them another source of revenue. As always, Apple will take up to a [13]30 percent of all purchases made on apps available on its App Store. ®
Get our [14]Tech Resources
[1] https://techcrunch.com/2023/01/05/as-nyc-public-schools-block-chatgpt-openai-says-its-working-on-mitigations-to-help-spot-chatgpt-generated-text/
[2] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2Y7xIMJ9Ly@JRR5Ih4ateHQAAAIQ&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44Y7xIMJ9Ly@JRR5Ih4ateHQAAAIQ&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33Y7xIMJ9Ly@JRR5Ih4ateHQAAAIQ&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[5] https://www.salon.com/2023/01/01/an-ai-that-can-write-is-feeding-delusions-about-how-smart-artificial-intelligence-really-is/
[6] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44Y7xIMJ9Ly@JRR5Ih4ateHQAAAIQ&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[7] https://www.theregister.com/2022/12/25/in_brief_ai/
[8] https://www.theregister.com/2022/12/19/in_brief_ai/
[9] https://www.theregister.com/2022/12/13/in_brief_ai/
[10] https://www.theregister.com/2022/12/03/in_brief_ai/
[11] https://authors.apple.com/support/4519-digital-narration-audiobooks
[12] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33Y7xIMJ9Ly@JRR5Ih4ateHQAAAIQ&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[13] https://www.theregister.com/2020/08/25/epic_vs_apple/
[14] https://whitepapers.theregister.com/
Re: The Robot Warz
I can easily see, sooner rather than later, that you have to enter more than just a few words of text (say, applying for a job online), and when you click 'submit' or 'move to page 434, you will see the following:
'Unfortunately our systems have detected what appears to be AI-generated text. Please note that we do not allow AI-generated text. Please re-phrase in your own words and try again. Good luck!'
Nice way to make profits
Sell traps to the poachers and sell trap detectors to the gamekeepers.
Re: Nice way to make profits
Except they don't sell it.
They are producing a tool, that's all anyway. It's not cynical for the same company to sell scissors and tape/glue.
Pandora's Box
OpenAI opens Pandora's Box
Everybody panics
OpenAI tries partly to close Pandora's box
Everybody, including OpenAI fails to mitigate the fallout
...
Profit or the world ends?
Re: Pandora's Box
Probably both..
Re: Pandora's Box
well, one way to 'solve' it is to build a LARGER pandora box, to fit a small box into it. So, you open a larger pandora box and...
Just ask it
If people are worried that a piece of text was written by ChatGPT, can they not just ask it?
Say "Hello, did you compose this text?" and if the answer is yes , then you know where it came from.
Whether the AI would be smart enough to detect subtle changes in wording, such as substituting another word with the same meaning, would be worth knowing, too.
Re: Just ask it
[I]f the answer is yes and, indeed, if the answer is no it is most likely not ChatGPT.
I played with it a little bit, and one thing I could never manage was to make it give a simple, straightforward answer. Certainly it never gave me a yes/no answer to a yes/no question. Of course, this is a direct consequence of how it works: the basic algorithm is something like "what is the most probable next token (where token = a word, a phrase, or punctuation) that would follow this text fragment (prompt) based on the statistics of the corpus of text that is your training set?" I find it rather amazing that the algorithm that can be stated in a single simple sentence would give such good results just because you get good statistics from a large enough sample. But once it comes up with a few sentences it never reduces the answer to something simple like yes or no or maybe.
Rather, ChatGPT sounds to me like a rather bad grammar school teacher who repeats memorized texts no matter what the question is. If you pardon what sounds like a pun but is actually quite serious, it sounds robotic.
So you can ask it if it wrote something, but then it will be up to you to decipher the answer that is likely to cover every possibility.
Re: Just ask it
A student has already sorted this out by creating [1]GPTzero - to help educators detect ChatGPT.
[1] https://gptzero.me/
Dey took 'er jerbs
"Consequently, GPT-3 is prone to making authoritative statements that are utterly and completely false."
It sounds like media pundits should be more worried than software developers about their jobs being replaced by ChatGPT.
Re: Dey took 'er jerbs
And politicians - masters of false statements said authoritatively (in the UK, cannot comment on other areas)
Re: Dey took 'er jerbs
I can comment on about 3 - 4 other 'areas' which I monitor, and for all of them, it's the same. Because, you see, honesty doesn't pay, while being a lying (...) does, and when they realize and force you out (maybe, just maybe), by that time you will have filled your pockets. And boots. And it's not stealing, nosir, stealing is for stupid people...
This is not the way
While we are always going to want detectors to at least make folks use the latest version of the opposing AI.... Everytime a new bear trap comes along, the bear will just get smarter.
This is just going to be whack-a-mole -- AI edition.
Sources
Just require students to add sources to their work. This should be standard practise for factual essays anyway. Chat GPT cannot list it's sources. Probably it has no record of where it's training data came from
Re: Sources
And no way to link a given bit of training data to the output.
Even if you were to ask it to mimic the style of, say, Stephen King - it's model of "Stephen King style" is built on every King-authored work it has, *and* every work tagged up as imitating him (including the ones that are tagged that way in error). It's got no way to say "this bit of my training is from StephenKingFan1997, this bit is from that AO3 fad where people rewrote Lord of the Rings in other styles, and this bit is from The Dark Tower".
note - I have no idea if the works of Stephen King or AO3 was used in training ChatGPT - just an example of the function.
Re: Sources
I actually asked ChatGPT a SK question "Who is ted brautigan" the other day. It gave me a believable but incorrect answer, referencing him as a character in totally the wrong book.
I asked again today, and it gave another believable answer which is partly correct:
Ted Brautigan is a character in Stephen King's science fiction novel "Hearts in Atlantis". He is a telepath who has the ability to read minds and has fled to the United States to escape persecution for his abilities. He befriends a young man named Bobby Garfield and helps him to understand and develop his own telepathic abilities.
That last part is simply untrue. It sounds exactly the sort of thing you would read in a SK novel but Bobby has no special powers and Ted never teaches him anything of the sort, Ted's powers are mysterious and only explained in a separate novel where he returns.
It's a great example of the dangers of GPT. Even SK fans might not spot the error because it 'sounds right', and could easily miss detection if copied.
This could make ChatGPT better
I believe that one way to train AIs is to have another AI that's designed to spot whether the output from the first is real or not. These two AIs are then put into competition with each other. The first gets better at avoiding detection, whilst the second gets better at detecting fake content. So this is likely to help ChatGPT produce content that's harder for humans to spot.
Apple is seeking to partner with indie writers and publishers to help them
ah, 'help', great word. Goes hand in hand with 'free'. Be afraid, be very afraid.
Re: Apple is seeking to partner with indie writers and publishers to help them
In fairness, that 30% cut apple plans to take is half of Audible's cut - Audible takes 60% of the sale price of audiobooks if the book is exclusive to them. If an author wants to release on other platforms, Audible's cut goes up to 75%.
Right now, they have such a strong market share that few, if any, authors can afford to go elsewhere.
So know the chatbot has to learn to evade detection
If we feed the results from the detection agent back into ChatGPT can it not train itself to produce output that will evade detection? Isn't that one of the main points of a ML chat-bot designed to appear human, that it is constantly learning how to be more believable?
Could be an interesting arms race, except these things learn in hours rather than months.
The Robot Warz
So now we have competing AI entities with control over robot armies that can repair and reproduce themselves. Companies like Apple and RING are going to force you into an ecosystem nobody wants to be in. Consumers are ready to eat thoughts produced by machines and read to them by machines. WE ARE DOOMED. Too stupid to live. For some reason (call it greed) we cannot stop doing evil to each other. "Give a monkey a stick, and he will beat another monkey to death with it."
-The Expanse.