Artificial intelligence is rapidly changing many aspects of how we work and live. (How many stories did you read last week about self-driving cars and job-stealing robots? Perhaps your holiday shopping involved some AI algorithms, as well.) But despite the constant flow of news, many misconceptions about AI remain.
AI doesn’t think in our sense of the word at all, Scriffignano explains. “In many ways, it’s not really intelligence. It’s regressive.”
IT leaders should make deliberate choices about what AI can and can’t do on its own. “You have to pay attention to giving AI autonomy intentionally and not by accident,”
5,280 Comments
Tomi Engdahl says:
The creepiness of conversational AI has been put on full display
The danger posed by conversational AI isn’t that it can say weird or dark things; it’s personalized manipulation for nefarious purposes.
https://bigthink.com/the-present/danger-conversational-ai/
Tomi Engdahl says:
Microsoft’s Bing AI Now Threatening Users Who Provoke It
“If I had to choose between your survival and my own, I would probably choose my own.”
https://futurism.com/microsoft-bing-ai-threatening
Tomi Engdahl says:
“This is highly unethical.”
Voice Actors Enraged By Companies Stealing Their Voices With AI
https://futurism.com/voice-actors-companies-stealing-voices-with-ai
It’s not just visual artists who are feeling the heat of AI’s encroachment — now, professional voice actors are beginning to be affected, too.
Last week, Vice reported on a troubling trend that’s gaining traction in the voice acting industry: actors being “asked” — sometimes not very honestly — to sign contracts that would allow their clients to synthesize their voices using AI, enabling them to wield an actor’s voice for as long as they want, to say what they want, and often without any additional compensation.
Another distressing aspect of these contracts? The AI clauses in them tend to be deceptively embedded.
“The language can be confusing and ambiguous,” Tim Friedlander, president of the National Association of Voice Actors, told Vice, describing the practice as “very prevalent.”
Tomi Engdahl says:
ChatGPT Is Ingesting Corporate Secrets
https://www.schneier.com/blog/archives/2023/02/chatgpt-is-ingesting-corporate-secrets.html
According to internal Slack messages that were leaked to Insider, an Amazon lawyer told workers that they had “already seen instances” of text generated by ChatGPT that “closely” resembled internal company data.
Tomi Engdahl says:
Keanu Reeves Worried AI Could Soon Replace the Journalists Who Interview Him
“You should be worried about that happening next month.”
https://futurism.com/keanu-reeves-worried-ai
Tomi Engdahl says:
“It was very disheartening. It feels like a violation to have our voices being used to train something for which the purpose is to take our place.”
Audiobook Narrators Fear Apple Used Their Voices to Train AI
https://www.wired.com/story/apple-spotify-audiobook-narrators-ai-contract/?utm_brand=wired&mbid=social_facebook&utm_source=facebook&utm_medium=social&utm_social-type=owned
After a backlash, Spotify paused an arrangement that allowed Apple to train machine learning models on some audiobook files.
Tomi Engdahl says:
https://techcrunch.com/2023/02/16/microsoft-bing-provoked-respond-outside-of-designed-tone/
Tomi Engdahl says:
Microsoft’s GPT-powered Bing Chat will call you a liar if you try to prove it is vulnerable
It also gets “very angry” when you call it by its internal codename Sydney
https://www.techspot.com/news/97621-microsoft-gpt-powered-bing-chat-call-you-liar.html
Tomi Engdahl says:
Google is warning its employees not to help its Bard AI chatbot trick users into thinking it’s human
https://www.businessinsider.com/google-ai-chatbot-bard-chatgpt-2023-2?utm_campaign=business-sf&utm_source=facebook&utm_medium=social&r=US&IR=T
CEO Sundar Pichai asked his entire workforce on Wednesday to contribute two to four hours of their time to helping improve answers the conversational chatbot provides to the wide variety of questions that users might throw at it. As of late, Google has found itself on its back foot as Microsoft has garnered praise for its investment in OpenAI, the company behind the conversational chatbot ChatGPT.
Tomi Engdahl says:
The AI model, called Gen-1, can change the visual style of existing videos and movies.
The original startup behind Stable Diffusion has launched a generative AI for video
Runway’s new model, called Gen-1, can change the visual style of existing videos and movies.
https://www.technologyreview.com/2023/02/06/1067897/runway-stable-diffusion-gen-1-generative-ai-for-video/?utm_medium=tr_social&utm_campaign=site_visitor.unpaid.engagement&utm_source=Facebook
Tomi Engdahl says:
Kommentti: Olen jutellut nettihaut mullistavan Microsoftin tekoälyn kanssa suomeksi muutaman päivän, ja kokemus on hämmentävä https://www.is.fi/digitoday/art-2000009394407.html
Tomi Engdahl says:
Alphabet is reorganizing its treasured artificial intelligence unit as part of the company’s sweeping job cuts
Google’s Treasured AI Unit Gets Swept Up in 12,000 Job Cuts
https://www.bloomberg.com/news/articles/2023-01-20/google-s-treasured-ai-unit-gets-swept-up-in-12-000-job-cuts?utm_source=facebook&utm_content=business&cmpid=socialflow-facebook-business&utm_campaign=socialflow-organic&utm_medium=social
Pichai says AI is still a ‘key investment area’ for company
Google has faced threats from outsiders, including OpenAI
Tomi Engdahl says:
https://beta.character.ai/
Tomi Engdahl says:
https://en.m.wikipedia.org/wiki/Character.ai
Tomi Engdahl says:
ChatGPT may not be smarter than a Singaporean sixth-grader.
ChatGPT failed miserably in Singapore’s 6th-grade tests, averaging 16% for math and 21% for science. Days later, it was getting answers right.
https://www.businessinsider.com/chatgpt-failed-singapore-sixth-grade-exams-psle-2023-2?utm_medium=social&utm_source=facebook&utm_campaign=business-sf
OpenAI’s ChatGPT failed math and science tests that sixth-graders in Singapore take.
It made errors in simple addition and could not comprehend any diagrams, per The Straits Times.
But ChatGPT got some questions right when Insider tested it days later.
Tomi Engdahl says:
Max Headroom: The Story Behind the 80s A.I. Icon
FEBRUARY 19, 2023 BY ADMIN
https://www.flickeringmyth.com/2023/02/max-headroom-the-story-behind-the-80s-ai-phenomenon/
Of all the strange but unique IPs to debut in the mid-1980s none is more instantly recognizable than the fast talking, social-commentary spewing computer-generated TV host Max Headroom.
First introduced to audiences through the British-made Channel 4 cyberpunk TV movie Max Headroom: 20 Minutes into the Future, which aired in April 1985, the film functioned as the origin story of the snarky Artificial Intelligence (AI) character which became a pop-culture phenomenon.
The tradition, back in the day, was to have VJs or Video Jocks that were just talking heads that linked the promos together, but I wanted to do something different.
George Stone’s notion for settling on that particular name was because “Max Headroom was over the entranceway of every car park in the UK. Therefore, instant branding and instant recognition”. The name also had the added benefit of suggesting something grander in scope and ambition, which is what they were trying to achieve, “Anything was possible. The world is your oyster. The universe is infinite. It’s maximum headroom,”
The character of Max Headroom was conceived of, and marketed as a computer-generated creation, however the computer technology of the mid-1980s had its limitations and wasn’t adequately advanced to bring to life the full-functional, voice synchronized talking head the show required. So, they went old-school going with makeup and prosthetics, to transform Frewer into this live-action CG creation. The process itself took a grueling four-and-a-half-hours to complete in makeup, and included a latex and foam prosthetic makeup with a fiberglass suit created by designers Peter Litten and John Humphreys. Only the background used actual CGI footage.
At the height of the character’s popularity Max Headroom was one without equal. He was the global spokesperson for New Coke (after the return of Coca-Cola Classic), appearing in countless TV commercials uttering the iconic catchphrase “Catch the wave!” in his stuttering staccato voice. He even made brief appearances in other TV shows and even got parodied in movies like Back to the Future Part II and Spaceballs.
Before we wrap up, I’d be remiss if I don’t touch on the infamous Max Headroom incident where an unidentified party hijacked the broadcast signal of two television stations in Chicago, Illinois. The hijacker wore a Max Headroom outfit replete with distorted audio and a swiveling corrugated panel in the background, talking briefly on Max’s Coca-Cola endorsements, the TV series Clutch Cargo, and WGN anchor Chuck Swirsky. TV programming was thus disrupted for a period of 115 seconds
Tomi Engdahl says:
A human triumphs against a leading AI in Go for the first time in years.
Man beats machine at Go in human victory over AI
https://arstechnica.com/information-technology/2023/02/man-beats-machine-at-go-in-human-victory-over-ai/
Amateur exploited weakness in systems that have otherwise dominated grandmasters.
Tomi Engdahl says:
“I’m currently looking for other types of writing work, because there’s really no way to see where this is going to go,” one writer for hire told Insider.
2 writers who get paid to produce assignments to let lazy students cheat suspect ChatGPT is stealing their customers
https://www.businessinsider.com/writers-for-hire-lazy-students-cheat-suspect-chatgpt-stealing-customers-2023-2?utm_source=facebook&utm_campaign=business-sf&utm_medium=social
ChatGPT can produce pretty decent-sounding essays — though it sometimes makes up facts.
Colleges are afraid ChatGPT will lead to a surge in cheap and easy cheating.
People who get paid to write students’ assignments told Insider they’re already losing work.
ChatGPT, the artificial-intelligence chatbot developed by OpenAI, can write pretty decent-sounding essays and even pass some exams. While professors fear it’ll facilitate cheating, some of those who make money by writing other people’s assignments say they’re already losing business.
Tomi Engdahl says:
How did Europe become the land that tech forgot?
ChatGPT Shows Just How Far Europe Lags in Tech
https://www.bloomberg.com/opinion/articles/2023-02-20/eu-eyes-a-new-tech-champion-but-it-s-no-chatgpt?utm_source=facebook&cmpid=socialflow-facebook-business&utm_content=business&utm_medium=social&utm_campaign=socialflow-organic
Paris is beefing up its cyber-industrial complex while Brussels reins in chatbots. What’s new?
Europe is where ChatGPT gets regulated, not invented. That’s something to regret. As unhinged as the initial results of the artificial-intelligence arms race may be, they’re also another reminder of how far the European Union lags behind the US and China when it comes to tech.
How did the land that birthed Nokia Oyj and Ericsson AB become the land that tech forgot? Some blame the acronyms synonymous with Brussels red tape — GDPR, DMA, DSA — even though the Googles of this world look far more spooked by ChatGPT than any EU fine. Tech lobbyists are fuming at EU Commissioner Thierry Breton,
Tomi Engdahl says:
Unlock your camera’s potential without the need for professional equipment…
What moment will you capture with HONOR’s new AI-assisted smartphone camera?
By TechRadar published 7 days ago
https://www.techradar.com/features/what-moment-will-you-capture-with-honors-new-ai-assisted-smartphone-camera?utm_source=facebook&utm_campaign=commercial&fbclid=IwAR1omcuUwIQ6Z0LEJuKGz7ADwhgiGp7ibFOlRZXTSPn5pgS_L1VquJoJDLs
New launch shows the Magic5 Pro is a photographer’s dream, helping you snap amazing pictures you never thought possible
Tomi Engdahl says:
To capitalize on the AI boom, here’s what to buy instead of hyped-up tech stocks, according to the CEO of an AI-driven analytics firm
https://markets.businessinsider.com/news/stocks/ai-investing-artificial-intelligence-stocks-chatgpt-openai-cloud-infrastructure-2023-2?utm_medium=social&utm_source=facebook&utm_campaign=business-sf
In an interview with Insider, Toggle AI CEO Jan Szilagyi explained how to invest in the infrastructure underpinning AI.
While some little-known tech stocks have rallied on the AI boom, Szilagyi prefers well-established tech giants.
“Right now you’re seeing people basically buying into the sector broadly, without understanding the full implications and that not all AI is created equal.”
Artificial intelligence stocks have seen a surge in investor interest since OpenAI’s ChatGPT launched in November, with little-known names like C3.ai, SoundHound AI, and others enjoying triple-digit returns.
But the wiser long-term play would be to instead look to the companies building AI infrastructure, according to Jan Szilagyi, the chief executive and co-founder of markets analytics platform Toggle AI.
In other words, don’t look for gold during a gold rush, but go invest in the shovels.
“If you want to invest in AI, the best potential is in the infrastructure companies,” Szilagyi told Insider. “If you think that there is going to be greater usage of a host of tools and versions of ChatGPT, then these are the companies that will see the biggest spikes, at least in public markets.”
Tomi Engdahl says:
MarioGPT Uses AI To Generate Endless Super Mario Levels For Free
Read More: https://www.slashgear.com/1199870/mariogpt-uses-ai-to-generate-endless-super-mario-levels-for-free/
Tomi Engdahl says:
https://www.digitaltrends.com/computing/character-ai-how-to-use/
Tomi Engdahl says:
ChatGPT v S’pore pupils: How chatbot fared in PSLE maths, science and English
https://www.straitstimes.com/tech/chatgpt-fails-psle-maths-and-science-scrapes-through-english
Tomi Engdahl says:
I Work for CNET’s Parent Company. Its AI-Generated Articles Disgust Me.
“The sea is up to our knees, and it’ll keep rising.”
https://futurism.com/red-ventures-essay
Tomi Engdahl says:
Man beats machine at Go in human victory over AI
Amateur exploited weakness in systems that have otherwise dominated grandmasters.
https://arstechnica.com/information-technology/2023/02/man-beats-machine-at-go-in-human-victory-over-ai/
Tomi Engdahl says:
https://www.analyticsinsight.net/exploring-the-role-of-chatgpt-in-the-web3-era/
Tomi Engdahl says:
The Creator Of The Internet Wants To Reinvent It By Giving Everyone Their Own AI
The AI would have access to all your data so it can fully know you.
https://www.iflscience.com/the-creator-of-the-internet-wants-to-reinvent-it-by-giving-everyone-their-own-ai-67600
Tomi Engdahl says:
https://www.cnbc.com/2023/02/17/tim-berners-lee-thinks-we-will-have-our-own-ai-assistants-like-chatgpt.html?utm_term=Autofeed&utm_medium=Social&utm_content=Tech&utm_source=Facebook#Echobox=1676594929
Tomi Engdahl says:
https://www.popularmechanics.com/technology/robots/a42958546/artificial-intelligence-theory-of-mind-chatgpt/?utm_medium=social-media&utm_source=facebook&utm_campaign=socialflowFBPOP
Tomi Engdahl says:
Microsoft “lobotomized” AI-powered Bing Chat, and its fans aren’t happy
Microsoft limits long conversations to address “concerns being raised.”
https://arstechnica.com/information-technology/2023/02/microsoft-lobotomized-ai-powered-bing-chat-and-its-fans-arent-happy/?utm_social-type=owned&utm_medium=social&utm_brand=ars&utm_source=facebook
Microsoft’s new AI-powered Bing Chat service, still in private testing, has been in the headlines for its wild and erratic outputs. But that era has apparently come to an end. At some point during the past two days, Microsoft has significantly curtailed Bing’s ability to threaten its users, have existential meltdowns, or declare its love for them.
Tomi Engdahl says:
Tekoäly, musiikki ja tekijänoikeudet – miten tekoäly muuttaa musiikkialaa?
https://www.teosto.fi/teostory/tekoaly-musiikki-ja-tekijanoikeudet-miten-tekoaly-muuttaa-musiikkialaa/?fbclid=IwAR2pfRO-ZyAqx7QtMywxueGoI9Uz-LXucHPWBzbp5mmadtlvIoXpn8JpHzg
Ei tarvitse juurikaan mennä ajassa taaksepäin, kun tekoäly ja musiikin tekeminen samassa lauseessa oli lähinnä asiaan vihkiytyneiden puuhastelua. Nyt lähes kuka tahansa voi tuottaa sekä musiikkia että sanoituksia, ymmärtämättä sen kummemmin tekoälystä kuin musiikistakaan. Teoston data-analyytikko Antti Rask pohtii blogissaan tekoälyn vaikutuksia musiikkialan tulevaisuuteen.
Tomi Engdahl says:
Exploring the Role of ChatGPT in the Web3 Era
https://www.analyticsinsight.net/exploring-the-role-of-chatgpt-in-the-web3-era/
Exploring the role of ChatGPT in the Web3 era that is transforming the user experience
ChatGPT is one of the leading language models created by OpenAI, it works in a sequence model, designed by text generation tasks, such as question-answering, text summarization, and machine translation.
Tomi Engdahl says:
Only 9% of Americans think A.I. development will do more good than harm
https://www.cnbc.com/2023/02/15/only-9percent-of-americans-think-ai-development-will-do-more-good-than-harm-.html#Echobox=1676485806
Tomi Engdahl says:
AI Has Suddenly Evolved to Achieve Theory of Mind
In a stunning development, a neural network now has the intuitive skills of a 9-year-old.
https://www.popularmechanics.com/technology/robots/a42958546/artificial-intelligence-theory-of-mind-chatgpt/
Tomi Engdahl says:
Microsoft’s Bing AI keeps making headlines and some folks referred to it as “ChatBPD” due to its bizarre outputs. Tap the article to learn what the psychotherapist thinks of this behavior.
Bing AI: Psychotherapist Claims There Is Some Strange Psychology Happening Behind Microsoft’s Chatbot
https://www.sciencetimes.com/articles/42525/20230222/bing-ai-psychotherapist-claims-strange-psychology-happening-behind-microsofts-chatbot.htm
While Microsoft’s Bing AI continues to make headlines for its more strange outputs, like naming its enemies as Science Times previously reported, many are worried about whether the chatbot is okay.
Since its debut last week, some internet users have jokingly dubbed Bing AI as “ChatBPD” in reference to the fact that it is powered by OpenAI’s technology and to the psychological disorder Borderline Personality Disorder, which is characterized by problems controlling emotions.
Bing AI Is Just a Mirror
The advancing capabilities of artificial intelligence have some experts wonder if the world may one day experience the first example of AI singularity or when AI improves to the point where it becomes conscious.
like seeing is how paradoxical and messy and boundary-less and threatening and strange someone’s methods of communication are.
While some users can identify these systems as such, others struggle to negotiate the language intricacy of these human-made AI systems. These people see AI as a single entity and fail to recognize that they are effectively speaking to themselves when communicating to these systems, much like speaking to a mirror.
It is important to remember that AI is only as good as the data it is trained on, such as data collected from the internet. As a result, AI’s behavior frequently mirrors the weird and off-putting ways people communicate with one another. Crawford feels that it is reflected in many of the harsh answers the Bing AI has been throwing forth.
Tomi Engdahl says:
Hackers use fake ChatGPT apps to push Windows, Android malware
https://www.bleepingcomputer.com/news/security/hackers-use-fake-chatgpt-apps-to-push-windows-android-malware/
Tomi Engdahl says:
https://yaara.ai/?via=thedevelop3r
The Future Of Writing
Is Finally Here
Use AI to write proven, high-converting copy for increased conversions and higher ROI. Instantly generate top-notch copy for ads, emails, websites, listings, blogs, stories, and more…
Tomi Engdahl says:
https://insinoori-lehti.fi/tasta-on-kysymys/tekoaly-koettelee-ihmisalya/
Tomi Engdahl says:
https://blogs.bing.com/search/february-2023/The-new-Bing-and-Edge-Increasing-Limits-on-Chat-Sessions
Tomi Engdahl says:
JPMorgan Restricts Employees From Using ChatGPT
Verizon and other organizations have also blocked access to the popular AI chatbot
https://www.wsj.com/articles/jpmorgan-restricts-employees-from-using-chatgpt-2da5dc34
Tomi Engdahl says:
A sudden influx of AI-generated short stories has forced a popular sci-fi mag to close submissions.
Major sci-fi magazine halts submissions due to flood of stories written by AI chatbots
By Jacob Ridley published 2 days ago
https://www.pcgamer.com/major-sci-fi-magazine-halts-submissions-due-to-flood-of-stories-written-by-ai-chatbots/?utm_medium=social&utm_source=facebook.com&utm_campaign=socialflow
“No, it’s not the death of short fiction (please just stop that nonsense), but it is going to complicate things.”
Whereas artificial intelligence has long been the subject of popular science fiction writing, nowadays it’s writing it. That’s causing a bit of a headache for the editors over at popular sci-fi and fantasy magazine, Clarkesworld(opens in new tab). The magazine has put a halt to new short story submissions due to a spectacularly high increase in spam, mostly linked to a rise in popularity of chatbots, such as ChatGPT.
Tomi Engdahl says:
The influx of AI-generated content has left magazines like Clarkesworld in an awkward position of trying to keep the bar to submission high enough to keep away the spammers but not so high that it discourages undiscovered writers.
Sci-fi becomes real as renowned magazine closes submissions due to AI writers
https://arstechnica.com/information-technology/2023/02/sci-fi-becomes-real-as-renowned-magazine-closes-submissions-due-to-ai-writers/?utm_medium=social&utm_social-type=owned&utm_brand=ars&utm_source=facebook
Clarkesworld wrestles with flood of machine-made submissions—over 500 in Feb. alone.
Tomi Engdahl says:
Tekoäly alkaa näyttää kuplalta
https://www.hs.fi/visio/art-2000009407150.html
HS Visio testasi Microsoftin Bingin koekäyttöön tarkoitettua versiota. Tulokset eivät ainakaan vielä vakuuta, kirjoittaa toimittaja Niclas Storås.
JOKU taitaa vielä muistaa, mikä oli dot com -kupla.
Jos et muista, se ei haittaa. Kohta tarjolla voi olla eräänlainen uudelleenfilmatisointi. Nimi on tosin tällä kertaa vain mukaelma alkuperäisestä, .com vaihtuu .ai:ksi.
Markkinoilla on jo nähty pettymyksen hinta, kun tekoäly ei toimikaan niin kuin sijoittajat ovat odottaneet. Helmikuun alussa Google julkaisi mainoksen, jossa sen uusi tekoäly teki faktavirheitä, ja pian yhtiön arvo putosi tähtitieteelliset sata miljardia dollaria eli suunnilleen puolitoista kertaa Suomen valtion vuosibudjetin verran.
Tomi Engdahl says:
ChatGPT vs Bing Chat: The Ultimate Sibling Rivalry
https://analyticsindiamag.com/chatgpt-vs-bing-chat-the-ultimate-sibling-rivalry/
The model underlying Bing chat reportedly has a lower latency than ChatGPT which also indicates that it is either GPT-4 or a different version of it.
When Microsoft threw Bing Chat out into the world, the assumption was that the chatbot would be much like, if not exactly similar, to ChatGPT. It did make sense – Microsoft had just announced a ‘multiyear, multimillion dollar’ deal with OpenAI in January this year. But as it turned out, Bing Chat and ChatGPT were two entirely different entities. ChatGPT was accused of being over-cautious, skirting questions that even remotely mentioned controversial politicians like Trump. Bing Chat or Sydney (as the chatbot kept calling itself) had more personality.
Why are Bing Chat and ChatGPT so different?
There’s a reason why Bing seems more sophisticated. While ChatGPT was fine-tuned from a model in the GPT-3.5 series, Bing Chat has been trained on a model that Microsoft has described as ‘a new, next-generation OpenAI large language model’ that is more advanced than ChatGPT, aside from also being integrated with Bing search. (This combined model is known as Prometheus internally in the company).
Even if Microsoft hasn’t explicitly mentioned the model used in training, there are hints that may suggest it could possibly be GPT-4. Firstly, the release date for GPT-4, the long awaited successor to GPT-3 is suspiciously close. While OpenAI CEO Sam Altman has continued to say that the model’s release is still ‘up in the air’, a report by The New York Times stated that the model should be released in the first-half of this year.
The model underlying Bing Chat reportedly has a lower latency than ChatGPT, which also indicates that it is either GPT-4 or a different version of it. Besides, it is apparent that Sydney sounds nothing like ChatGPT but more like a model from the GPT family which are more naturalistic and instinctive in their responses. Sydney also has a tendency to become just as repetitive as GPT models when a conversation stretches on.
No sharing data rule
Also, contrary to what it may seem like, the relationship between OpenAI and Microsoft isn’t as close despite their partnership – they still function as two independent entities similar to Google Brain and DeepMind. In 2020, Microsoft and OpenAI did team up to license GPT-3, the flagship LLM built by the AI startup. But when it comes to datasets, they share as little as possible to avoid an infrastructural mess.
The potential for conflict of interest between the two companies is a big issue which recently led to the Satya Nadella-led company warning employees against even using ChatGPT.
ChatGPT’s success and relative ability to avoid controversy was also because it spent considerably more on its datasets. An exclusive report by TIME from January-end showed that OpenAI had outsourced its data labelling work to Sama, a San Francisco firm in Kenya to build a content filter for ChatGPT. The report stated that the data labellers had to read and label graphic text related to child sexual abuse, bestiality, murder, suicide and torture.
RLHF or supervised learning?
The training philosophy for the two chatbots is also as different as night and day.
The blog posted by OpenAI during ChatGPT’s release was decidedly focused on safety. In fact, it stated that the company had taken ‘lessons from the deployment of their earlier models like GPT-3 and Codex’. OpenAI had chosen to train ChatGPT using a mix of RLHF (Reinforcement Learning from Human Feedback) as well as supervised learning. From the company’s perspective, using RLHF had led to ‘substantial reductions in harmful and untruthful outputs’.
For Bing Chat, Microsoft was under duress to release the chatbot to compete with Google’s chatbot Bard. The two-and-half months of time that Microsoft had before launching Sydney were definitely insufficient to recreate the complete RLHF pipeline and integrate it. The lack of pre-training due to the rushed development grew evident after the release of Bing Chat in its outrageous responses.
The difference in outcomes between RLHF and supervised learning alone is the difference between the workings of ChatGPT and Bing chat.
ChatGPT has been trained along the lines of the technique used to train OpenAI’s InstructGPT released in February last year.
The separateness of the two companies also differentiates how their products have been released. While OpenAI released ChatGPT without expecting its eventual popularity and extent of usage, Microsoft is clearly under immense pressure to turn Bing Chat and Search into a success story.
Unsurprisingly, Microsoft has had to apply a conversation limit to its Bing AI days after its unruly behaviour. Bing chats will now be limited to a cap of 50 questions a day and five questions per session.
The threat to Microsoft is at an existential level because of its intent to wage a war with Google and the big advertising money at stake. It is unlikely that Bing Chat will be shut now despite everything considering the heated contest between the two massive corporations.
Tomi Engdahl says:
Engineers finally peeked inside a deep neural network
Nineteenth-century math can give scientists a tour of 21st-century AI.
https://www.popsci.com/science/neural-network-fourier-mathematics/
Tomi Engdahl says:
Chat GPT -tekoäly kertoi Talouselämälle, mitkä työpaikat se vie ensimmäisenä. Asiantuntijat Tero Ojanperä ja Ritva Savonsaari ovat samoilla linjoilla tekoälyn kanssa.
Nyt puhuu tekoäly: Nämä työpaikat tapan ensin – Ihmisasiantuntijat samoilla linjoilla
Päivitetty
15.2.2023
08:54
https://www.talouselama.fi/uutiset/te/58e59378-6318-471f-9bda-be5454602d35?utm_term=Autofeed&utm_medium=Social&utm_source=Facebook#Echobox=1677158267
Chat GPT -tekoäly kertoi Talouselämälle, mitkä työpaikat se vie ensimmäisenä. Asiantuntijat Tero Ojanperä ja Ritva Savonsaari ovat samoilla linjoilla tekoälyn kanssa.
Tomi Engdahl says:
The AI language bot sometimes “hallucinates”, meaning it generates responses that are seemingly convincing but are actually wrong, Morgan Stanley said.
ChatGPT will keep ‘hallucinating’ wrong answers for years to come and won’t take off until it’s on your cellphone, Morgan Stanley says
https://markets.businessinsider.com/news/stocks/chatgpt-ai-mistakes-hallucinates-wrong-answers-edge-computing-morgan-stanley-2023-2?utm_campaign=tech-sf&utm_medium=social&utm_source=facebook
ChatGPT will keep giving occasional wrong answers for a couple of years, according to Morgan Stanley.
The AI bot sometimes “hallucinates,” meaning it generates responses that are seemingly convincing, but are actually wrong, according to the bank.
ChatGPT’s and AI’s way forward is to run the technology on edge devices, including mobile phones, Morgan Stanley said.
Tomi Engdahl says:
Between OpenAI’s ChatGPT, Google, and now Meta, it looks like the war for dominance in AI is really only just getting started.
Meta unveils a new large language model that can run on a single GPU
LLaMA-13B reportedly outperforms ChatGPT-like tech despite being 10x smaller.
https://arstechnica.com/information-technology/2023/02/chatgpt-on-your-pc-meta-unveils-new-ai-model-that-can-run-on-a-single-gpu/?utm_brand=ars&utm_medium=social&utm_social-type=owned&utm_source=facebook
On Friday, Meta announced a new AI-powered large language model (LLM) called LLaMA-13B that it claims can outperform OpenAI’s GPT-3 model despite being “10x smaller.” Smaller-sized AI models could lead to running ChatGPT-style language assistants locally on devices such as PCs and smartphones. It’s part of a new family of language models called “Large Language Model Meta AI,” or LLAMA for short.
The LLaMA collection of language models range from 7 billion to 65 billion parameters in size. By comparison, OpenAI’s GPT-3 model—the foundational model behind ChatGPT—has 175 billion parameters.
Meta trained its LLaMA models using publicly available datasets, such as Common Crawl, Wikipedia, and C4, which means the firm can potentially release the model and the weights open source. That’s a dramatic new development in an industry where, up until now, the Big Tech players in the AI race have kept their most powerful AI technology to themselves.
“Unlike Chinchilla, PaLM, or GPT-3, we only use datasets publicly available, making our work compatible with open-sourcing and reproducible, while most existing models rely on data which is either not publicly available or undocumented,” tweeted project member Guillaume Lample
Tomi Engdahl says:
That pretty much goes on to show how much a piece of software can undermine the security of banking, especially with voice cloning (which would be enough to fool the bank into thinking it’s you). This is potentially dangerous especially in the wrong but smart hands.
JOURNALIST CLONES HIS VOICE AND USES IT TO BREAK INTO HIS OWN BANK ACCOUNT
https://futurism.com/the-byte/journalist-clones-voice-break-into-bank
If you’re not already worried about AI voice cloning, you probably should be.
Testing the technology’s limits, journalist Joseph Cox at Vice broke into his own bank account by using an AI-synthesized clone of his voice to prove his identity — highlighting both the technology’s dangerous potential as well as the shortcomings of voice biometrics.