#chatgpt
guter 🧵zur Warnung vor "tödlicher KI" ausgerechnet durch KI CEOs wie Sam Altman (OpenAI - #ChatGPT). Es ist vor allem eins: ein gigantisches Ablenkungsmanöver u der Versuch, die Regulierung aktuell schon existierender #Risiken von #KI zu vermeiden. ⬇️
RT @LuizaJarovsky: 🔥Regulate us, but not really
I've just left an in-person event with Sam Altman and Ilya Sutskever (OpenAI's CEO & Chief Scientist) at Tel Aviv University, and these were my impressions: https://t.co/3I75uv1ovF
"#ChatGPT ist Autokorrektur auf Steroiden." Katharina @netzwerkerin Zweig liefert (erneut) einen großartigen Vergleich und blickt in den Maschinenraum der #KI-Entwicklung. #rp23
5 AI tools for summarizing a research paper - Unlock the power of AI tools to extract key insights and condense... - https://cointelegraph.com/news/5-ai-tools-for-summarizing-a-research-paper #researchpapersummarization #condensinginformation. #informationextraction #keyinsights #quillbot #scispacy #aitools #chatgpt
Aktuell findet die Ringvorlesung “ChatGPT und Co.” an der Uni Regensburg statt. Heute mit dem Thema “Will ChatGPT take my job? Spotting stochastic parrots in the wild”. https://www.uni-regensburg.de/sprache-literatur-kultur/information-medien-sprache-kultur/institut/veranstaltungen/index.html #dh #dhregensburg #digitalhumanities #chatgpt #chatbot #ai #ki
"We've seen ChatGPT generate URLs, references and even code libraries and functions that do not actually exist." https://www.infosecurity-magazine.com/news/chatgpt-spreads-malicious-packages/
@rikefranke #ChatGPT increases my work productivity a lot when it comes to developing and coding models. It's a source of inspiration, typing assistant, sparring partner, and helps troubleshooting.
My Bluesky/Mastodon/Posthaven cross-poster is coming along nicely! Thanks also to #chatGPT of course ;)
Can share code if anyone is interested.
Wie immer sehr hörenswert der Research Library Podcast (RLP) der RLP034. Digital Humanities unter KI Bedingungen. Gespräch mit Thomas Wallnig.
Episode webpage: https://rlp.univie.ac.at/?podcast=rlp034-digital-humanities-unter-ki-bedingungen-gespraech-mit-thomas-wallnig
Darin wird auch kurz die Rolle von #Bibliotheken und #OpenAccess für aktuelle LLM angesprochen: sie bieten qualitativ hochwertige Information und machen die neuen Werkzeuge damit besser.
5 ChatGPT Tricks To Enrich Your Coding Journey With Intuitive & Efficient Learning
https://medium.com/@ai_millennia/a-guide-to-code-better-using-chatgpt-f53dc6c1f834
#MachineLearning #Math #DataScience #100DaysOfMLCode #DataAnalytics #DataScientist #100DaysOfCode #Statistics #ChatGPT #ArtificialIntelligence #ai #OpenAI #coding
Heute um 18 Uhr an der
Uni Regenburg und via Zoom: "Will #ChatGPT take my job? Spotting stochastic parrots in the wild" – Vortrag von AlgorithmWatch-Reporter Nicolas Kayser-Bril im Rahmen der interdisziplinären Veranstaltungsreihe "ChatGPT & Co."
➡️ Programm & Zoom-Link: https://www.uni-regensburg.de/sprache-literatur-kultur/information-medien-sprache-kultur/institut/veranstaltungen/index.html
#Cryptocurrencies #adoption #AI #Business #ChatGPT Meta’s Zuckerberg grilled by senators over ‘leak’ of LLaMA AI model: Two United States senators have questioned Meta chief executive Mark Zuckerberg over the tech giant’s “leaked” artificial intelligence model LLaMA which they claim is potentially “dangerous” and could be used for “criminal tasks.”
In a June 6 letter, U.S. Senators Richard Blumenthal and Josh Hawley criticized Zuckerberg’s decision to open source… https://btcnews.biz/cryptocurrencies/meta-s-zuckerberg-grilled-by-senators-over-leak-of-llama-ai-model/?utm_source=dlvr.it&utm_medium=mastodon
Tim Cook: “What people are now talking about are these Large Language Models. I think they have great promise. I do think that it’s so important to be very deliberate and very thoughtful in the development and the deployment of these because they can be so powerful that you worry about things like bias, things like misinformation, maybe worse in some cases.”
#ClimateCrisis
#ChatGPT
#cartoon
Honestly the response to climate crisis across many nations are just as silly as this.
TL;DR: #ChatGPT fails as stuff because we're expecting it to do things it wasn't designed for.
It wasn't designed to get math right. It wasn't designed to get *anything* right - it was designed to make something that *looks like* it's right.
Again: It wasn't built for correctness, but for *the appearance* of correctness.
Not that #AI vendors are all that eager to tell you this. They are, after all, trying to build a #business on your misunderstanding.
Don't waste time juggling hundreds of prompts. All you need are these 5 powerhouse prompts for #ChatGPT!
Learn more 👉 https://ngmi.gumroad.com/l/nobsprompts
At this point, #ChatGPT is like a man - quite likely going to be wrong, but it will be very confident while being entirely wrong.
Funny how much #ChatGPT thinks adding “low protein” in front of food makes it so, like low protein tofu and soy sauce.
Yes, let me have my low protein chicken seasoned with low protein glass shards 👍👍👍
Twintigste bloemlezing over ChatGPT (en aanverwante technologieën)
We zijn alweer bij de twintigste bloemlezing over ChatGPT en aanverwante technologieën aanbeland. Met dit keer acht samenvattingen en ongeveer vijftig links naar Nederlandstalige bronnen of bronnen waar ik niet aan toe kom om ze samen te vatten. https://www.te-learning.nl/blog/twintigste-bloemlezing-over-chatgpt-en-aanverwante-technologieen/
#artificialintelligence #generatieveai #chatgpt #digitalegeletterdheid #didactiek #onderwijs #kritiek #edutoot
Listening to Sam Altman on Lex Fridman podcast and I like the dude. I don't think he's right about everything but he puts in the work that's for sure. I'm also satisfied by his reasoning about why #openai went for-profit + non-profit control. Maybe he's not that bad of a leader for this space after all.
Great listen even if you don't care about #chatgpt or #agi
https://open.spotify.com/episode/6rAOusZcsuNtCv8mefmwND?si=HWcWjvRpRg-5JuhjzP5Ijg
Mittwoch: OpenAI geht nicht an die Börse, Milliarden für französische Chipfabrik
Kein Börsengang für OpenAI + Globalfoundries-Fab in Frankreich + Nord-Stream-Sabotage vorab bekannt + Rushkoff-Tipps für Superreiche + Analyse der Apple-Brille
#Apple #AugmentedReality #ChatGPT #Erdgas #Globalfoundries #KünstlicheIntelligenz #VirtualReality #Subventionen #UkraineKrieg #news
The Rise of OpenAI: Advancing Artificial Intelligence for the Benefit of Humanity
OpenAI is a research organization that ... #Artificialintelligence #ChatGPT #ChatGPTPlus #ComputerVision #DALLE #ElonMusk #FutureofAI #GenerativePretrainedTransformer #GPT #NaturalLanguageProcessing #OpenAI #OpenAIsGPTmodel #Robotics #SamAltman #Videogamebots https://www.knewtoday.net/the-rise-of-openai-advancing-artificial-intelligence-for-the-benefit-of-humanity/?feed_id=12259&_unique_id=648007ee84a83
If you’re an advertiser scrambling to figure out how to “leverage AI”, please don’t forget the tons of machine learning your programmatic platforms have been using for like a decade.
Tell your boss you’re ahead of the curve! #advertising #chatgpt https://en.m.wikipedia.org/wiki/Infinite_monkey_theorem
as many people have already pointed out, #chatGPT is a very good #bullshitter
>#ChatGPT woulld not eliminate humans to solve the #climatecrisis :)
Pity! 😜
OpenAI-Chef nicht interessiert am Börsengang seiner KI-Firma
Laut Sam Altman würden Anleger manche Entscheidungen bei der Entwicklung einer Superintelligenz skeptisch sehen. Ein Rückzug aus der EU ist nicht vorgesehen.
#Börse #ChatGPT #ElonMusk #EU #KünstlicheIntelligenz #Microsoft #news
If #chatGPT isn't good for writing regexes for you, what good is it at all?
Cool use of #ChatGPT
News Minimalist — Only significant news https://www.newsminimalist.com/
@bykriscampbell I've noticed that Chat-GPT uses certain words way more than any normal human - especially, "seamlessly", as in "X implementation seamlessly aligns with Y manifesto". Ironic that that word in particular makes it feel rather.. seamful. Reminds me of this paper: "Seamful XAI: Operationalizing Seamful Design in Explainable AI"
ChatGPT in your terminal. Let's have a strong ally by your side.
I asked #ChatGPT to write Tim Cook's keynote product announcement at the 2033 WWDC. #AppleVisionPro
...and now... One. More. Thing.
It's totally bizarre to me that #Cyan released a new #Myst-like game, it's in #VR, and much of it was built by #AI and people are *mad* about it.
The fuck is wrong with you ingrateful non-contributing zeroes? It's a small studio, it's a gorgeous game, and #generativeAI is a miracle of computer science.
Y'all need to go hug your waifu pillows and shut up.
#aiArt #aiArtwork #chatGPT #gpt3 #llm #videoGames #gaming
https://kotaku.com/myst-firmament-cyan-studios-ai-machine-writing-art-pc-1850505094
So lately I'll be reading something -- maybe a blog post, maybe a press release, etc. And I come across a sentence that seems perfectly acceptable but just immediately strikes me the wrong way. I feel it viscerally, like a punch in the gut.
When I've been able to ask the author about the sentence in question, it turns out they've had help from chat GPT and the sentence is AI-generated.
Has this happened to anyone else?? I think I must have a highly attuned AI-dar. 😂
Geo-hallucination? A geojson polygon file created by #chatgpt (24 May version) for the city limits of #victoria BC (red) vs city opendata of the same https://opendata.victoria.ca/datasets/VicMap::city-boundary/explore?location=48.440729%2C-123.359375%2C11.90 (yellow) #DHSI
“My aunt who I live with has a parrot 🦜…” 👉 https://vm.tiktok.com/ZTRooeUnq/
There’s nothing new about #ChatGPT ! 😂
Informatiklehrer: „Wenn jetzt ChatGPT Version 4.0 so viel besser arbeitet, wie arbeitet dann eine #ChatGPT 5.0 oder 6.0? Wie sollen Lehrer an den Schulen später feststellen, ob die Lösungen von den Schülern gemacht worden sind oder von irgendwelchen Künstlichen Intelligenzen? Das finde ich schon ein wenig beunruhigend.“ #FediLZ #LernenmitKI
„Note 5+, Note 4-, setzen: Erst vor wenigen Monaten ist die #KünstlicheIntelligenz am bayerischen Abitur gescheitert. Hinfallen, Krone richten, weitergehen? Ja, offenbar hat #ChatGPT Nachhilfe von der Entwicklerfirma OpenAI bekommen.“ https://ardaudiothek.de/episode/11km-der-tagesschau-podcast/ki-zwischen-abi-und-apokalypse/tagesschau/12733639/ (30min) #LernenmitKI #FediLZ #abitur
Make Use Of: Snapchat’s My AI or ChatGPT: Which One Should You Use? https://www.makeuseof.com/snapchat-my-ai-or-chatgpt/ #Tech #MakeUseOf #TechNews #IT via @morganeogerbc #ArtificialIntelligence #Internet #Snapchat #ChatGPT #Chatbot
Messed around with the ChatGPT api to see if it could simulate data in the style of a Stroop experiment, complete with RT data. #cognition #chatgpt
https://crumplab.com/blog/771_GPT_Stroop/
It did something close enough, need to spend a bit more time on this tomorrow
Thinking about the possibilities of ChatGPT bots completing JsPsych experiments...
#ChatGPT has made me an 11x engineer.
PrivateGPT - Running "ChatGPT" offline on local documents https://dev.to/codepo8/privategpt-running-chatgpt-offline-on-local-documents-5b17
By Chris Heilmann codepo8@toot.cafe #ChatGPT
How to create a private ChatGPT that interacts with your local documents
"By using a local language model and vector database, you can maintain control over your data and ensure privacy while still having access to powerful language processing capabilities. The process may require some technical expertise, but there are many resources available online to help you get started."
I think this is the general way things are headed.
https://bdtechtalks.com/2023/06/01/create-privategpt-local-llm/
Bestes Beispiel für „schön-schaurig“: Auf Stage 1 das „Ende von Twitter“ mit @dennishorn und @gavinkarlmeier in Jeopardy-Form zelebrieren. Mit @mrtoto festgestellt, dass der Preis „Mastodon for Dummies“ bestimmt von #ChatGPT geschrieben wurde. #rp23
"Semantic #search", going from literal (token-based) matches to getting closer to the meaning of your text, picture,... at #JNation
https://xeraa.net/talks/elasticsearch-vector-hybrid-search/
PS: Yes, it also mentions #ChatGPT 😉
#ChatGPT: Navigating the rising financial crime landscape in the digital age #AI #cybersecurity #infosec https://betanews.com/2023/06/06/chatgpt-rising-financial-crime-landscape/ @betanews
The sort of people who would do a bad job* using #ChatGPT are also the sort of people that were about to do a bad job without ChatGPT.
If you took away Stackoverflow, the sort of people who ask lazy 0 effort questions will still be putting 0 effort into their task.
* By bad job I mean, the people who'd ask & copy-paste without reading or checking the answer
It does occur to me that the smartest people on the planet are rarely the richest or the most powerful.
I would imagine the same issues would face a general AI.
💬 Deep-dive into the AI Hardware of ChatGPT
➥ High Yield
#ChatGPT #HPC #NVIDIA #CarbonFootprint
https://m.youtube.com/watch?v=4q9-yf1eU8c&feature=youtu.be
#ChatGPT brings an interesting problem: how do we determine whether its results are correct? The only way I can see is crowd-sourcing model inputs, google captcha style. “Is this helpful? Click yes or no”
Getting this right could allow a new global corporation to emerge. With the state of global capitalism, we the people have a choice: do we want this? Wrong data on a massive, coordinated scale will prevent it.
Embrace your inner Boaty McBoatface, for this is gonna continue to be a wild ride.
Manipulationsgefahr: EU-Kommission fordert rasch Kennzeichnung von KI-Inhalten
Mit ChatGPT & Co. produzierte Inhalte sollen für Nutzer klar erkennbar sein. Kommissarin Věra Jourová spricht Maschinen das Recht auf freie Meinungsäußerung ab.
After effectively debunking all sorts of crypto and blockchain related nonsense, Amy Castor and @davidgerard have now turned their attention to the current #AI hype.
Lots of good stuff, including some much needed background to the alarmist warnings on AI risk by various AI luminaries.
Their characterization of #ChatGPT as "a stupendously scaled-up autocomplete" is spot-on as well.
https://davidgerard.co.uk/blockchain/2023/06/03/crypto-collapse-get-in-loser-were-pivoting-to-ai/
‘Talking About Large Language Models’, Murray Shanahan. Great paper!
‘The more adept LLMs become at mimicking human language, the more vulnerable we become to anthropomorphism, to seeing the systems in which they are embedded as more human-like than they really are. This trend is amplified by the natural tendency to use philosophically loaded terms, such as "knows", "believes", and "thinks", when describing these systems.‘ https://arxiv.org/abs/2212.03551 #tech #ai #llms #chatgpt
ChatGPT is not a technological singularity. ~ Daniel Lemire. https://cacm.acm.org/blogs/blog-cacm/273606-chatgpt-is-not-a-technological-singularity/fulltext #ChatGPT
Ich arbeite im Toursimus. Und als ich vor 5 Jahren bei einer Konferenz die These augestellt habe, dass in 10 Jahren Leute nicht zwingend zu ihrer Traumdestination reisen, sondern auch einfach mal ihre #VR-Brille aufsetzen und durch von einer #KI generierte Landschaften wandern/skifahren, haben alle mitleidig gelächelt. Jetzt haben wir #ChatGPT und demnächst kommt #VisionPro ...
Bertrand Meyer: "AI does not help Programmers" (#ChatGPT)
https://m-cacm.acm.org/blogs/blog-cacm/273577-ai-does-not-help-programmers/fulltext
Here are the first and last versions of the function, and it's final response. I never gave it any hints or directions other than the very first prompt telling it what the function should do.
My next prompts were either, "Consider your answer carefully for any flaws, mistakes, or wording that could be improved. Do you want to revise your answer? If so, go ahead and give a revised version. If not, why not?"
Or, "Should you revise any part of this new answer?"
![function convertTelephoneNumbersToLinks() {
// Get all the text content within the webpage
var bodyText = document.body.innerHTML;
// Regular expression to match telephone numbers
var telephoneRegex = /(\+\d{1,3}[-.\s]?)?\(?\d{3}\)?[-.\s]?\d{3}[-.\s]?\d{4}/g;
// Replace all telephone numbers with "tel:" links
var convertedText = bodyText.replace(telephoneRegex, function(match) {
return '<a href="tel:' + match.replace(/[-.\s]/g, '') + '">' + match + '</a>';
});
// Update the webpage with the converted text
document.body.innerHTML = convertedText;
}
// Call the function to convert telephone numbers to links
convertTelephoneNumbersToLinks();](https://assets.toot.cafe/cache/media_attachments/files/110/494/549/553/502/891/small/d5c951d00b5716ec.jpg)
![function convertTelephoneNumbersToLinks(node = document.body) {
var telephoneRegex = /(\+\d{1,3}[-.\s]?)?\(?\d{3}\)?[-.\s]?\d{3}[-.\s]?\d{4}/g;
if (node.nodeType === Node.TEXT_NODE) {
var text = node.nodeValue;
if (telephoneRegex.test(text)) {
var fragment = document.createDocumentFragment();
var lastIndex = 0;
var match;
telephoneRegex.lastIndex = 0; // Reset the lastIndex
while ((match = telephoneRegex.exec(text)) !== null) {
var prefix = text.substring(lastIndex, match.index);
var number = match[0];
if (prefix) {
fragment.appendChild(document.createTextNode(prefix));
}
var link = document.createElement('a');
link.href = 'tel:' + number.replace(/[-.\s()]/g, '');
link.textContent = number;
fragment.appendChild(link);
lastIndex = match.index + number.length;
}
if (lastIndex < text.length) {
fragment.appendChild(document.createTextNode(text.substring(lastIndex)));
}
node.parentNode.replaceChild(fragment, node);
}
} else {
for (var i = 0; i < node.childNodes.length; i++) {
convertTelephoneNumbersToLinks(node.childNodes[i]);
}
}
}
convertTelephoneNumbersToLinks();](https://assets.toot.cafe/cache/media_attachments/files/110/494/549/653/894/607/small/737f83a14f390035.jpg)
I asked #ChatGPT to write #JS to turn phone numbers in a page to "tel:" links. It gave a function that technically works, but also has bad side effects. Asked it to revise its answer if necessary. It gave me a new fn that was better. Kept asking it if it should revise, it kept improving until eventually it said, no, its answer didn't need revising. The final fn is impressive and has no issues I can tell. Really strange how it can improve its work. I'm impressed and feel a bit uneasy. #AI #LLM
😴 Tired: Yet another list of 10 AI tools that 1000x your life
🔥 Wired: A list of the 10 overhyped AI tools that failed miserably
🎯 Challenge: Can anyone create that list to save us?
🫡 I promise I'll boost it 🚀
USA: Allein im Mai 3900 Stellen wegen Künstlicher Intelligenz abgebaut
Allein im vergangenen Monat wurden laut einer US-Firma in den Vereinigten Staaten fast 4000 Stellen mit der Begründung "Künstliche Intelligenz" gestrichen.
I find it's much faster to obtain advice from #ChatGPT (4) on using Python's argparse module than trying to piece together the required code from the module's sometimes incomprehensible reference documentation.
"ChatGPT, automating being confidently, completely wrong since 2023™"
Star Trek: The Next Generation S1E16: When The Bough Breaks. In which the crew encounter the Aldeans, who have abdicated all their decisions to a computer “the Custodian” many hundreds of years ago. The Aldeans do not understand (and cannot maintain) the Custodian, nor any of its rationale in decision making. Yet they entrust it with all their decisions. They are on the brink of extinction, despite their technical prowess.
Poignant satire of chatGPT. First aired… 1991. #StarTrek #chatgpt
🧵
1/ English is the predominant language of the Internet. So what's different about ChatGPT squishing non-English languages farther into the margins?
Key points from this piece, my reflections, and one key recommendation for those in Responsible AI:
#AI #chatgpt #responsibleAI #academia #llm #mastodon
https://www.wired.com/story/chatgpt-non-english-languages-ai-revolution/
“Users speak of #ChatGPT as ‘hallucinating’ wrong answers — #LargeLanguageModels make stuff up and present it as fact when they don’t know the answer. But any answers that happen to be correct were ‘hallucinated’ in the same way.” — @davidgerard, https://davidgerard.co.uk/blockchain/2023/06/03/crypto-collapse-get-in-loser-were-pivoting-to-ai/
“#AI” “#ArtificialIntelligence” #GPT #GPT3 #GPT4 “#OpenAI” #LLM
I block #crypto skeptics, not because they are critical of something I'm invested in, but because I look at the rest of their posts and they are so generally full of negativity and derision for all sorts of progress.
They frame their personalities by what they are *against* and I find that toxic.
#bitcoin #btc #cryptocurrency #blockchain #ai #chatGPT #lmm #llms #gpt #aiart
Just saw someone on #Reddit using #ChatGPT to answer a user's question — they didn't disclose it but it was painfully obvious.
Very on-brand, as a large portion of Reddit's user base consists of people who are confidently wrong about everything.
Still, I miss the times before #AI, when being wrong at least required some effort.
#Asstodon Artificial Intelligence experiment:
I asked #ChatGPT, "Are donkeys more intelligent than robots?" and the answer was, "No, donkeys are not more intelligent than robots. Donkeys are animals and possess their own form of intelligence, adapted for survival in their natural environment." (Full response in ALT text.)
Then I asked Morris donkey, "Are donkeys more intelligent than robots?" and he replied, "I'm bored with my toys. What I really would like is a drone."
1/2 General Purpose AI is a huge double #threat to #humanity, but people are only concentrating on the 2nd part of the threat, the #Singularity, which is something we won't be able to do anything about anyway because it's mostly in the hands of #BadActors who won't listen to #regulation at all.
We will likely be in a world where #NorthKorea will be the #AI equivalent of being a primary nuclear power, for example.
Is it possible this is the greatest, most humane use of #ChatGPT ever? Detoxifying the internet?
(via @Sarahp’s @Flipboard magazine “This Week in Apps”)
Diese Woche war das hier besonders aufsehenerregend. Eine drastische Warnung vor #KI von denen, die daran arbeiten. Aber es ist halt wieder nichts Konkretes. Was soll man denn nun daraus machen? Haben die zu viel #ScienceFiction gelesen, fallen die auf ihren Hype rein oder müssen wir uns wirklich Sorgen machen?
"So gefährlich wie Pandemien und Atomkrieg": Nächste Warnung aus KI-Branche
#Atomwaffen #ChatGPT #GeoffreyHinton #KünstlicheIntelligenz #OpenAI #Pandemie
Sieben Tipps, damit Sie souverän bei KI-Themen mitreden können
Alle reden über KI. Wenn Sie unsicher sind, wovon zum Teufel die Leute reden, machen Sie sich keine Sorgen. Wir haben ein paar Leitlinien für Sie.
#ChatGPT #Infotech #KünstlicheIntelligenz #Roboter #Spracherkennung #Suchmaschine #news
What I learned at GlueCon about Generative AI - this blog post took me a week to finish writing, and it’s already out of date. Developments in generative AI are moving faster than most people who are part of it can keep up with...
https://adrianco.medium.com/what-i-learned-at-gluecon-2023-tipping-points-and-generative-ai-898cd5219040 #ChatGPT #generativeai #GlueCon #innovation #tippingpoint
Yes, #ChatGPT et al can do incredible seeming and very new things, is largely proprietary and opaque to most people, and is being hacked and experimented with at runaway pace far ahead of regulation.
But to jump from that to "existential threat", you have to clearly distinguish it from technological changes that have come before. I'm not sure it is clearly distinct.
They mentioned that the trouble with "#AI" is that it's "unknowable", it's a black box. Meaning that there's randomness to it. 2/