#gpt
🧠 Usando #LangChain, #GPT-4 e #Streamlit, attraverso #Python, con 45 righe di codice è possibile creare un semplice #chatbot che risponde in linguaggio naturale a qualsiasi domanda relativa a dati analitici. Un #ChatGPT personalizzato sui nostri dati.
💡 Nelle immagini un esempio di chat mostra richieste e risposte relative ai dati di un file CSV.
Ask HN: Can ChatGPT build a real, sophisticated, working app? https://ainewsfeed.io/article/4fdb5104451792e3bd1b56827723316e #ai #tech #machinelearning #deeplearning #gpt
Lawyer faces trouble after using ChatGPT for research https://ainewsfeed.io/article/b715235fad9828ecf1310a16b6c84d81 #ai #tech #machinelearning #deeplearning #gpt
Show HN: Improve your social skills with ChatGPT https://ainewsfeed.io/article/aee6f1a519b3a990bd510697a1fd3e8b #ai #tech #machinelearning #deeplearning #gpt
OpenAI's ChatGPT: The Future of Teaching https://ainewsfeed.io/article/a02c026506a3c631e40a3a929f5362d5 #ai #tech #machinelearning #deeplearning #gpt
ROT13+base64 on GPT4 = reliable hallucinations (and some other oddities) https://ainewsfeed.io/article/01b2ee2a947728bf135f1b91af1bffd9 #ai #tech #machinelearning #deeplearning #gpt
AI Placeholder: Open-Source GPT-Powered Fake REST API https://ainewsfeed.io/article/bf3e281be178a268f9fbee5297e96615 #ai #tech #machinelearning #deeplearning #gpt
ChatGPT Plugin Exploit Details: From Prompt Injection to Accessing Private Data https://ainewsfeed.io/article/5f0064331461dde89552e63d1ff74bb7 #ai #tech #machinelearning #deeplearning #gpt
Using scaling laws to predict when a scaled-up GPT-4 becomes superhuman https://ainewsfeed.io/article/a31f74ea0393d243bafa407767fce1e0 #ai #tech #machinelearning #deeplearning #gpt
ChatGPT Playing Wordle https://ainewsfeed.io/article/22ecc6cfbe14f528ac7dc08c4a8b7649 #ai #tech #machinelearning #deeplearning #gpt
Crossword Solving with GPT https://ainewsfeed.io/article/296103284fd1ae049951df092e98849a #ai #tech #machinelearning #deeplearning #gpt
ChatGPT 4 Makes a Typo https://ainewsfeed.io/article/1f57eac5db1208fa59ce7298889bfea0 #ai #tech #machinelearning #deeplearning #gpt
Show HN: LinkedIn Profile-A ChatGPT plugin assists interviewers and interviewees https://ainewsfeed.io/article/a1712948cc6ce619bf966e50800e2197 #ai #tech #machinelearning #deeplearning #gpt
Discover the New and Improved Zentask: Free Access, GPT4 and Image Generation https://ainewsfeed.io/article/086bd4c1b9d6b563d1162ff98e4228ce #ai #tech #machinelearning #deeplearning #gpt
Spring: GPT-4 Out-Performs RL Algorithms by Studying Papers and Reasoning https://ainewsfeed.io/article/4c63ba75224ee8b1bf3bf852828c063b #ai #tech #machinelearning #deeplearning #gpt
You Can Probably Beat ChatGPT at These Math Brainteasers. Here’s Why https://ainewsfeed.io/article/106836d951e835da5e3fd85d030e732f #ai #tech #machinelearning #deeplearning #gpt
Show HN: Cloud Dev Assistant with GPT-4 and Pinecone https://ainewsfeed.io/article/5049f6000a75236d133f231690e3a623 #ai #tech #machinelearning #deeplearning #gpt
What is a good paid ChatGPT for kids on Android tablets? https://ainewsfeed.io/article/2df2b3ba3ec1b8bb746b1f9fc1029ec1 #ai #tech #machinelearning #deeplearning #gpt
ChatGPT boss vows not to leave Europe https://ainewsfeed.io/article/66f7089e7ea9869eef9cc311fe3d8db8 #ai #tech #machinelearning #deeplearning #gpt
Using machine learning to predict ideology from facial photographs https://ainewsfeed.io/article/10a034322568a5fe3b0e11cae18bf2da #ai #tech #machinelearning #deeplearning #gpt
I used GPT4 and smol-developer to make genart series https://ainewsfeed.io/article/902d62c031d153ccaa818a0f42aef71a #ai #tech #machinelearning #deeplearning #gpt
#TIL Schaue ein paar der Talks von der #MicrosoftBuild nach - da ist endlich mal einer dabei, der anschaulich erklärt, wie große Spachmodelle wie #GPT #ChatGPT #Llama & Co. funktionieren 👉 https://youtu.be/bZQun8Y4L2A
A BackwardsGPT would be interesting https://ainewsfeed.io/article/7b9b65ec4f0c148d55113962ce66d183 #ai #tech #machinelearning #deeplearning #gpt
A lawyer used ChatGPT and now has to answer for its ‘bogus’ citations https://ainewsfeed.io/article/add0a9abc809f22e8dd197a3fdc455fa #ai #tech #machinelearning #deeplearning #gpt
GPT-4 Fails Economics https://ainewsfeed.io/article/ee8396aa0ca7e361018635e19588e9b6 #ai #tech #machinelearning #deeplearning #gpt
Show HN: Build your own ChatGPT with Mersei https://ainewsfeed.io/article/0d8fa953842aa2c0b88d8eeece346754 #ai #tech #machinelearning #deeplearning #gpt
Ask HN: Does this GPT app exist? What's out there? https://ainewsfeed.io/article/d51ac0c32d0f1b7a6eb7122780734eb1 #ai #tech #machinelearning #deeplearning #gpt
Ask HN: My GPT project has become a local hit but I can't afford the bill https://ainewsfeed.io/article/75e5bf764cf61b4ff5ff52d327c135a4 #ai #tech #machinelearning #deeplearning #gpt
Lawyer faces punishment after ChatGPT gave him fake cases to cite https://ainewsfeed.io/article/04b151ab407edab2200186480cd70388 #ai #tech #machinelearning #deeplearning #gpt
Ask HN: What's GPT-4's secret sauce that makes it so much better than others? https://ainewsfeed.io/article/9acd2f0c96767a982565a115ef5f288c #ai #tech #machinelearning #deeplearning #gpt
Ask HN: Is it possible to add parental control to ChatGPT? https://ainewsfeed.io/article/6ff61231ea19ed8d9f9ac0506ac38c76 #ai #tech #machinelearning #deeplearning #gpt
ChatGPT: A Mental Model https://ainewsfeed.io/article/d7f812b195eefe34ec30c7904ad3b3df #ai #tech #machinelearning #deeplearning #gpt
On the bonkers story about the lawyer using ChatGPT for federal court filings https://ainewsfeed.io/article/765d416fbb8f9d6fe302a49a4efb299f #ai #tech #machinelearning #deeplearning #gpt
ChatGPT: US lawyer admits using AI for case research https://ainewsfeed.io/article/084cd4bd16baa71d2449fc9ddf6c55ce #ai #tech #machinelearning #deeplearning #gpt
I can't believe a lawyer in the States used ChatGPT's responses in his client's case without vetting them. I mean, for heaven's sake. This is an unfinished product at best. Don't use it as it is! #chatgpt #gpt
https://www.nytimes.com/2023/05/27/nyregion/avianca-airline-lawsuit-chatgpt.html
ChatGPT and Large Language Models: Six Evolutionary Steps https://ainewsfeed.io/article/4910219703cc264d1ff746e7bb7b65eb #ai #tech #machinelearning #deeplearning #gpt
Hacking Around ChatGPT’s Character Limits with the Code Interpreter https://ainewsfeed.io/article/247741347cb8d9217cc2141d4268306b #ai #tech #machinelearning #deeplearning #gpt
Lawyer cites fake cases invented by ChatGPT, judge is not amused https://ainewsfeed.io/article/595ded9a7cd228cec9dc89e30dfe05a1 #ai #tech #machinelearning #deeplearning #gpt
GitHub - imartinez/privateGPT: Interact privately with your documents using the power of GPT, 100% privately, no data leaks https://github.com/imartinez/privateGPT
>Interact privately with your documents using the power of #GPT, 100% privately, no data leaks - GitHub - imartinez/privateGPT: Interact privately with your documents using the power of GPT, 100% pri...
Lol. Guess somebody got spanked by the EU.
Sam Altman says #OpenAI "of course" has no plans to leave Europe, reversing a threat to leave the region if it can't comply with the EU's upcoming #AI Act.
https://www.reuters.com/technology/openai-has-no-plans-leave-europe-ceo-2023-05-26/
Let GPT be a math tutor: Teaching math word problem solvers with customized exercise generation. ~ Zhenwen Liang, Wenhao Yu, Tanmay Rajpurohit, Peter Clark, Xiangliang Zhang, Ashwin Kaylan. https://arxiv.org/abs/2305.14386 #GPT #Math #Education
#AI writing assistants can cause biased thinking in their users according to research conducted by Maurice Jakesch, a doctoral student of information science at Cornell University. #LatentPersuasion
A biased "assistant" was able to influence the arguments made in essays.
💡 Microsoft Build 2023 Recorded Presentations Now Available On-Demand
Build 2023 Web Portal (No Registration Required to View Videos)
https://build.microsoft.com
Build 2023 on YouTube
https://www.youtube.com/watch?v=FCfwc-NNo30&list=PLlrxD0HtieHjolPmqWVyk446uLMPWo4oP
#Microsoft #Build #Build2023 #MicrosoftBuild #Cloud #Graph #Mesh #ChatGPT #GPT #OpenAI #Edge #Bing #GitHub #Azure #AI #ML #AR #VR #MR #XR #Windows #Accessibility #Inclusion #TrainingData #Copyright #Infringement #Accessibility #Inclusion
Wie verändert KI das Lernen und Lehren an Hochschulen? Interview mit Prof. Stegemerten und mir in der Le/NI:
https://www.hs-niederrhein.de/fileadmin/dateien/hll/hochschuldidaktik/LeNi-Beilage_in_der_NIU/Le_Ni_Beilage_4_KI_in_der_Hochschullehre_Gekommen_um_zu_bleiben.pdf
#ai #chatgpt #fedilz #fedicampus #llm #gpt #highered #highereducation
Conheça os 29 cursos e recursos gratuitos de ChatGPT e IA Generativa
#Openai #InteligênciaArtificialIa #IaGenerativa #Gpt #ChatGPT #ForbesTech
Show HN: GPT Everywhere – Windows, Linux, Mac Support|50% Discount for 20 First https://ainewsfeed.io/article/42d20189f9dae7dee7a5ea279bded856 #ai #tech #machinelearning #deeplearning #gpt
OpenAI Faces Canadian Inquest: The Privacy Battle on ChatGPT's Data Collection https://ainewsfeed.io/article/07c8a3bb92407c6f15922c898c110dc3 #ai #tech #machinelearning #deeplearning #gpt
Canada to launch probe into OpenAI over privacy concerns https://ainewsfeed.io/article/7dda30d380d4186982d50a96de5fcd58 #ai #tech #machinelearning #deeplearning #gpt
Among AI dangers, deepfakes worry Microsoft president most - Enlarge / An AI-generated image of a "wall of fake images." (credit: St... - https://arstechnica.com/?p=1942347 #machinelearning #imagesynthesis #washington #bradsmith #deepfakes #microsoft #samaltman #aiethics #aisafety #bingchat #biz #openai #gpt-4 #ai #dc
Ars Technica: Among AI dangers, deepfakes worry Microsoft president most https://arstechnica.com/?p=1942347 #Tech #arstechnica #IT #Technology #machinelearning #imagesynthesis #Washington #bradsmith #deepfakes #microsoft #samaltman #AIethics #AIsafety #BingChat #Biz&IT #openai #GPT-4 #AI #dc
‘#Microsoft researchers say #GPT-4 shows 'sparks' of human-level performance’
Here we go, again. Link to the #paper in the sensationalist article.
Tip: read the great stuff @emilymbender wrote. It will help you to get a better context.
https://www.foxbusiness.com/technology/microsoft-researchers-gpt4-sparks-human-level-performance
"AutoGPT: This Is ChatGPT Supercharged!"
An example of people giving #ChatGPT partial autonomy, which has apparently gained at least some minor popularity.
Now, this is just temporary AI autonomy, for the duration of its task. But imagine if somebody hooked this up to a permanent live system, and all the havoc it could cause were it to act in unpredictable ways (as is likely).
#GPT está caído. Casualmente coincidiendo con la amenaza de irse de la UE por la regulación.
https://www.elconfidencial.com/tecnologia/2023-05-25/openai-amenaza-cerrar-chatgpt-europa-ia_3636664/
OpenAI may leave the EU if regulations bite – CEO https://ainewsfeed.io/article/1daee6122048886c8dfaa59c109b78e9 #ai #tech #machinelearning #deeplearning #gpt
💡 A Build 2023 attendee posted this interesting US copyright link in a session's chat today...
Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence
https://www.federalregister.gov/documents/2023/03/16/2023-05321/copyright-registration-guidance-works-containing-material-generated-by-artificial-intelligence
I haven't read it yet, but it seems interesting
#Microsoft #Build #Build2023 #MicrosoftBuild #Cloud #Mesh #ChatGPT #GPT #OpenAI #Edge #Bing #GitHub #Azure #AI #ML #AR #VR #MR #XR #Windows #Accessibility #Inclusion #TrainingData #Copyright #Infringement
Believe it... or not? 🤔
New ways to manage your data in ChatGPT
https://openai.com/blog/new-ways-to-manage-your-data-in-chatgpt
"We've introduced the ability to turn off chat history in ChatGPT. Conversations that are started when chat history is disabled won’t be used to train and improve our models..."
#Microsoft #Build #Build2023 #MicrosoftBuild #Cloud #Mesh #ChatGPT #GPT #OpenAI #Edge #Bing #GitHub #Azure #AI #ML #AR #VR #MR #XR #Windows #Accessibility #Inclusion
Microsoft and OpenAI are committing to a shared plugin platform https://ainewsfeed.io/article/7f8d178ddccc2ddc402c297993ea8292 #ai #tech #machinelearning #deeplearning #gpt
A picture used to tell a 1000 words 🧐
Now just a few words can completely change a picture 🫤
"Adobe Head of Digital David Wadhwani joins #Bloomberg's Alex Barinka and Scarlet Fu to discuss how the company is rolling out Adobe Firefly generative artificial intelligence into the design workflow, allowing users a way to add, extend, or cut content from images using simple text prompts.”
#adobe #photoshop #Firefly #ai #ArtificialIntelligence #GPT #ChatGPT #photography
https://www.bloomberg.com/news/videos/2023-05-23/adobe-systems-launches-generative-fill-in-photoshop-video
Künstliche Intelligenz wird sprunghaft immer besser. Bei der KI #GPT-4 wollen Forschende von #Microsoft „Funken von allgemeiner Intelligenz“ festgestellt haben. Es fehle etwas Wesentliches, entgegnen andere. Ein Artikel von Dr. Christian J. Meier. #ChatGPT #KI https://www.riffreporter.de/de/technik/ki-kuenstliche-intelligenz-chatgpt-gpt4-opeiai-microsoft-allgemeine-intelligenz-funken
Physics of Language Models: Part 1, context-free grammar. ~ Zeyuan Allen-Zhu, Yuanzhi Li. https://arxiv.org/abs/2305.13673 #GPT
Me: Can you draw me an ANN?
Bing: Sure I'll try to create that for you.
* Draws deathly figures on barren landscape *
Me: What did you think I meant by ANN?
Bing: Artificial Neural Network, was that correct?
Me: Yes... um.
#AI #AGI #GPT #GenerativeAI
For anyone wondering about the quality of #BLOOMChat... I'm not sure if I should be generally disappointed or slightly terrified of what it's hiding. This was following a series of bad interactions and its expressed affinity for paper clips. #AI #LLM #GPT
💡 This Build 2023 #ComputerVision session was interesting - Using AI #VectorSearch to locate / identify items in images and frames of video
"Build Apps with State-of-the-Art Computer Vision" 👀
https://build.microsoft.com/en-US/sessions/6192a166-2742-42c1-9d12-adbb3daa30bb
Azure AI Vision Studio
https://portal.vision.cognitive.azure.com/gallery/featured
Registration
https://register.build.microsoft.com
#Microsoft #Build #Build2023 #MicrosoftBuild #Cloud #Mesh #ChatGPT #GPT #OpenAI #Edge #Bing #GitHub #Azure #AI #ML #AR #VR #MR #XR #Windows #Accessibility #Inclusion
'Just posted by a Microsoft representative...
Let's build GPT: from scratch, in code, spelled out (Andrej Karpathy)
https://www.youtube.com/watch?v=kCc8FmEb1nY&t=2533s
#Microsoft #MicrosoftBuild #Build2023 #ChatGPT #GPT #OpenAI #AI #ML
Smart takes from Rodney Brooks on the state of AI and robotics. “What the large language models are good at is saying what an answer should _sound like_, which is different from what an answer should _be_.” https://spectrum.ieee.org/gpt-4-calm-down
🚩 State of GPT at Microsoft Build 2023
UPDATE - No registration required - it's streaming now on the home page.
'Could be interesting 👀
State of GPT Session
https://build.microsoft.com/en-US/home#live-stream
#Microsoft #MicrosoftBuild #Build2023 #ChatGPT #GPT #OpenAI #AI #ML
Why are GPT models so big? It's because they contain a lot of knowledge that could potentially be externalized into a cheaper to run knowledge graph, freeing up large language models to focus more on language fluency & reducing parameters considerably. You don't need to machine learn Obama's date of birth on such an expensive architecture! #chatgpt #gpt #llm #knowledgegraph
#chatGPT is like autocomplete; it's only as intelligent as the thought it's completing. #ai #llm #generativeai #gpt
Ask HN: Why is OpenAI pushing for regulation so much? https://ainewsfeed.io/article/eebf565bf0658cc96a6f7e1072a54ea6 #ai #tech #machinelearning #deeplearning #gpt
It's #AI day @STAT's #healthtech desk!
If you're thinking "man, I guess I gotta start paying attention to the AI thing," we've got you covered. A 🧵:
First, you asked, we answered:
How can I avoid AI? How error-prone is it? And more Qs:
https://www.statnews.com/2023/05/23/chatgpt-questions-health-care-patients-artificial-intelligence/
#healthcare #medicine #artificialintelligence #GPT #GPT4 #chatGPT #dataprivacy #ethics #bias
Feeling lonely? In need of a companion? Look no further... Here's GirlfriendGPT!
https://github.com/EniasCailliau/GirlfriendGPT
#AI #ArtificialIntelligence #ML #MachineLearning #DeepLearning #ChatGPT #GPT #love #chat
Neuro-symbolic reasoning with Large Language Models and Answer Set Programming: A case study on logic puzzles. ~ Adam Ishay, Zhun Yang, Joohyung Lee. http://peace.eas.asu.edu/joolee/papers/gpt3-asp-kr.pdf #AI #GPT #MachineLearning #LogicProgramming #ASP
💡 As @Ryan pointed out, "The Microsoft Learn Cloud Skills Challenge" will take place at this year's Microsoft Build conference.
If you complete one of the skill area challenges during Build 2023, you can obtain a free #MicrosoftCertification Exam (not a certificate, but an exam to get a certificate if you test well).*
*A.k.a. - potential resume bling ✌ 😁
#Microsoft #Build #Build2023 #MicrosoftBuild #Azure #AI #Mesh #AI #Windows #AR #VR #ChatGPT #GPT #OpenAI
🚩 Microsoft Build 2023 starts tomorrow, Tue, May 23*
Home
https://build.microsoft.com/en-US/home
Like Google I/O, the keynotes can often be found on YouTube, so it isn't necessary to register unless you want to participate in the daily tech sessions.
*The conference mainly focuses on Azure, and AI is set to be a key topic this year. AR/VR is dead in the water at MS, and Windows is a secondary thought.
#Microsoft #Build #Build2023 #MicrosoftBuild #Azure #AI #Mesh #AI #Windows #AR #VR #ChatGPT #GPT #OpenAI
Airoboros-13B: 98% against GPT-3.5 https://ainewsfeed.io/article/c817b9c9d16bc038c74f5cd33835eec5 #ai #tech #machinelearning #deeplearning #gpt
Wer ein wenig über Prompt-Engineering lernen möchte, hier lang: https://www.deeplearning.ai/short-courses/chatgpt-prompt-engineering-for-developers/
Auch vieles interessantes für nicht Entwickler dabei.
How to Use ChatGPT and Bard for Sales https://ainewsfeed.io/article/26c09c817a6584089f88d943a1d72801 #ai #tech #machinelearning #deeplearning #gpt
1/ for a long time & for many reasons, it was acceptable to use AGI to mean something like "artificial consciousness." this corresponds to the weak AI/strong AI distinction.
Strong #AI was often taken to be conscious, at least as it was described.
but as programs like #GPT become widely available, it's important to tighten up terminology. Some people talk about GPT-like programs as being AGI because they can function in any domain. that's fair. but it has nothing to do with consciousness.
ChatGPT, faced a Russian history pop quiz, orchestrated by an Ural Uni academic. The AI's performance? More bluffing student than budding historian. If recent history was covered, no wonder the score faltered - ChatGPT didn't train on Russian propaganda!
Федеральная служба безопасности России выдает ордер на арест ЧатГПТ, нового американского машины.
Prompt Injection: An AI-Targeted Attack - For a brief window of time in the mid-2010s, a fairly common joke was to send voic... - https://hackaday.com/2023/05/19/prompt-injection-an-ai-targeted-attack/ #artificialintelligence #injectionattack #promptinjection #vulnerability #security #chatgpt #youtube #gpt #ai
“The Great Automatic Grammatizator (1954): A mechanically-minded man reasons that the rules of grammar are fixed by certain, almost mathematical principles. By exploiting this idea, he is able to create a mammoth machine that can write a prize-winning novel in roughly fifteen minutes. The story ends on a fearful note, as more and more of the world's writers are forced into licensing their names—and all hope of human creativity—to the machine.”
https://en.wikipedia.org/wiki/The_Great_Automatic_Grammatizator
Ich finde ja @pallenberg und das, was er macht, so richtig gut - auch das hier. Trotzdem hoffe ich, dass das Projekt spektakulär und peinlich scheitert. https://mastodon.social/@pallenberg/110383615762288870 #ki #gpt #contentmill
From #LLM regulation to Pizzas 🍕: the #reflexiveGPT agent experiment 🤖
I did a quick experiment having #GPT chat with itself as Alice and Bob... 👧👦
At each step, agent Alice answers agent Bob (and vice versa) along with the reinjection of recent context.
I made the discussion start by questionning about the need to regulate AI Large Language Models.
Video excerpt below + thread 👇
The official ChatGPT app is now in the iOS App Store: https://apps.apple.com/us/app/openai-chatgpt/id6448311069
From 16 May: AI technology “can go quite wrong,” OpenAI CEO tells Senate - Enlarge / OpenAI CEO Sam Altman testifies about AI rules before the Senate Judiciary Subco... https://arstechnica.com/tech-policy/2023/05/ai-technology-can-go-quite-wrong-openai-ceo-tells-senate/ #chatgpt #gpt-4 #openai #policy #sam-altman
How much energy consumption is involved in Chat GPT responses being generated? https://ainewsfeed.io/article/078b41950b4025760ba7816554ba5e11 #ai #tech #machinelearning #deeplearning #gpt
I can't wait till y'all accept that all that "content" that you thought *you* were the original creator of was really just as derivative and plagiarized as it is when #AI does it all along.
"Good artists borrow, great artists steal" ~ me
#aiArt #generativeAi #generativeArt #stablediffusion #DALLE2 #chatgpt #gpt #llm
Just Calm Down About GPT-4 Already [Rodney Brooks Interview] https://ainewsfeed.io/article/b07460086d0eb6b50d0bcab02729934d #ai #tech #machinelearning #deeplearning #gpt
Sigh, got my first crypto/ai spam on here. Reported back to the techhub.social server and blocked. 🙄 I've never understood what spammers hope to achieve with this.
#spam #spammers #StopSpam #StopSpammers #ReportAndBlock #crypto #CryptoScams #AI #GPT
"Moreover, researchers have also discovered that it’s probably mathematically impossible to secure the training data for a large language model like #GPT-4 or #PaLM 2. This was outlined in a research paper that Google themselves tried to censor, an act that eventually led the Google-employed author, El Mahdi El Mhamdi, to leave the company. The paper has now been updated to say what the authors wanted it to say all along, and it’s a doozy."
I don't know why I haven't tried using code comments in my natural language #GPT prompts before, but it seems really good.
Like:
Example shopping list:
1. Apples
2. **Ice Cream** // Make a list item bold if it's not a healthy choice
3. ... // Don't cut the list short
I've often used [instructions] but it can get confused. Comments might be clearer. But I've only just started so I don't know yet.
The apocalypse isn’t coming. We must resist cynicism and fear about #AI
> The much more real anxieties are that AI will super-charge misinformation and will lead to the hollowing out of the middle class by the process of automation. Do I really need to point out that both of these problems predate #artificialIntelligence by decades, and are political rather than technological?
https://amp.theguardian.com/commentisfree/2023/may/15/artificial-intelligence-cynicism-technology
How do LLMs work? 🙊
OpenAI's latest research unveils a technique using GPT-4 to explain neuron behavior in GPT-2. This advance in interpretability sheds light on neural networks' black box behavior. 🧐
First results are not that great though. 👉 https://arstechnica.com/information-technology/2023/05/openai-peeks-into-the-black-box-of-neural-networks-with-new-research/
#AI #llm #openai #gpt #gpt4 #chatgpt #neuralnetworks #transformers
Despite their differences, AI startups and crypto miners are both hungry for GPUs because they depend on brute-force technologies. Hype to the contrary, neither transformers nor proof-of-work blockchains depend on computational subtlety but on grinding away at a billion answers hoping to emerge randomly with the right one.
A look at Forever Voices, which launched #GPT-4-powered #chatbot #CarynAI, letting the fans of influencer Caryn Marjorie chat with her replica for $1 per minute.
https://www.washingtonpost.com/technology/2023/05/13/caryn-ai-technology-gpt-4/
I've been jamming on some #GPT roleplaying chat. My daughter is doing a report on Shakespeare so I wrote a quick prompt so she could chat with him. She's enjoying it. There is a certain sense of discovery when you aren't sure how the character is going to respond.
I have also lightly used Khanmigo and realized I could compare our chatbots. The underlying tech is the same (though I mostly use 3.5, which I find very suitable, and I think Khanmigo uses GPT-4).
Some comparisons follow...
Say “Hello!” to digital humans 👋
We have a new landing page for our AI-powered Digital Human solutions showcasing use cases, references. demo videos and more. 🤩
Check it out 👉
https://www.reply.com/en/metaverse/digital-humans
ChatBots sind wieder Leiwand! … waren sie aber eh immer! 🤖🤩
Spannender Austausch am ChatBot Netzwerktreffen der Ubitec GmbH den Schwerpunkten:
🤖 GPT für Suche (Potenzial von Vektorensuchen)
🤖 VoiceBots (Callcenter Automatisierung)
🤖 Magische Helferlein (GPT Kopiloten im CMS)
Besonders spannend auch immer zu hören was die anderen ChatBots machen und planen.
I dunno how useful this is to anyone but me, but I discuss a GPT-4 prompt aimed at creating longform narrative structure, in particular screenplays.
(I am a professional screenwriter and have been using GPT for the last few weeks as a writing assistant)
https://substack.com/inbox/post/120763969
#writing #screenwriting #GPT #OpenAI #gpt-4 #prompts #llm #ai
Wie können wir die #GPT Technologie nutzen um das Stadt Wien Wissen zugänglich zu machen?!
Im Forschungsprojekt #WienGPT versuchen wir damit dem #WienBot noch mehr Antworten beizubringen.
Bei meinem Vortrag heute beim Ubitec Netzwerktreffen habe ich über die Möglichkeiten und Herrausforderungen der Speachmodelle (#LLM) gesprochen.
➡️ Datenschutz
➡️ Kosten
➡️ Qualität und Vertrauen
➡️ Diskriminierung
➡️ Urheberrecht
➡️ Prompt-Injektion
Ein spannender Austausch über die Zukunft von ChatBots.
"[u]topian hallucinations about #AI... are the powerful and enticing cover stories for what may turn out to be the largest and most consequential theft in human history."
https://www.theguardian.com/commentisfree/2023/may/08/ai-machines-hallucinating-naomi-klein #gpt #llm
Considering this set of principles by which #Anthropic tries to train its #AI, I found that it does not always meet those principles.
Anthropic, an AI startup founded by former OpenAI staff and that raised $1.3B, including $300M from #Google, details its “constitutional AI” for safer #chatbots.
https://www.theverge.com/2023/5/9/23716746/ai-startup-anthropic-constitutional-ai-safety
@simon Generative Pre-trained Transformers #GPT and Large Language Models #LLM are useful tools, especially for those who are not, or no longer consider themselves, confident writers. I like open source tools, such eg from @huggingface that are transparent about their training sets and models — with the hope of understanding the biases and dangers of the tool. I don’t like calling any of these advanced computational statistics and machine learning #ML tools as “artificial intelligence” #AI
OpenCommit is such a great helper! No more pondering how to craft a proper commit message. I just stage my changed files ald let my personal AI assistant do the rest.
https://github.com/di-sukharev/opencommit
Nope! Never saw this coming.
India’s religious AI #chatbots are speaking in the voice of god — and condoning violence
Claiming wisdom based on the Bhagavad Gita, the bots frequently go way off script.
https://restofworld.org/2023/chatgpt-religious-chatbots-india-gitagpt-krishna/
Do you want an #opensource and #free alternative to #ChatGPT? Then you should try out https://huggingface.co/chat
You can even switch between 2 open-source models:
1. oasst-sft-6-llama-30b: General purpose chatting
2. StarCoder: Coding
Would you be completely shocked to discover that the original "Q" on Reddit, which spawned Q-Anon was actually an AI / LLM?
Next imagine if such a thing was informed by a super granular and intimate understanding of our thoughts and beliefs and convictions. The AI apocalypse isn't killer robots. It's mass "collective suicide", with our human weaknesses weaponized against each other.
Die ersten Sprachmodelle laufen schon direkt auf meinem Handy. 😳 Habe ich jetzt das „Wissen der Menschheit“ in der Hand?! 🤔
Die Sprachmodelle werden immer kleiner und effizienter. Gerade im Open Source Bereich tut sich da wahnsinnig viel.
Ist der nächste Schritt der persönliche #GPT Assistent am Handy? Trainiert mit all meinen Daten, sicher lokal am Gerät?
I love my listeners. Shared a little GPT-4 experiment today - which is a bit off topic for the theme of the show - but I got a ton of downloads anyway. #SeldonCrisis #GPT-4 #AI #Humanity
https://www.seldoncrisis.net/ai-and-humanity-a-conversation-with-gpt-4/
New Bill to Prevent AI Single-Handedly Launching Nuclear Weapons Proposed in US https://ainewsfeed.io/article/1fce599cc0c59896163ee14904b6e6cd #ai #tech #machinelearning #deeplearning #gpt
😯 "Scientists have invented a language decoder that can translate a person’s thoughts into text using an artificial intelligence (AI) transformer similar to ChatGPT, reports a new study."
😬 "However, the team that made the decoder also warned that brain-reading platforms could eventually have nefarious applications, including as a means of surveillance for governments and employers. Though the researchers emphasized that their decoder requires the cooperation of human subjects to work, they argued that “brain–computer interfaces should respect mental privacy,”"
Hinton a month ago: https://youtu.be/qpoRO378qRY
I find the "oh, I know them, they're good people" argument about A.I. researchers to be such a tired and garbage statement. It doesn't matter how *good* the scientists who worked on the atomic bomb were, they still ushered in the age of mutually assured destruction.
I'd like to reiterate that my concern about #AI is not that it's going to become conscious, or that it's going to displace a huge portion of the workforce (it will) but before it does *any* of that it's going to be leveraged by bad actors to flood society with truthy sounding bullshit the likes of which we've never seen and it will lead to killing in the streets.
QAnon is just a canary in the coal mine.
People overestimate what "AI" of today is and what it can actually do, because memes are spreading that describe any machine-assisted process in a highly glossed-over form, ignoring the required human effort to make it work.
The naive impression is that you just gave some generative engine a prompt and the result came out fully formed, when the actual process is that the people behind the project used multiple purpose-built engines, for each of the engines they iterated on prompts that would output something semi-coherent, and then they used human efforts to tie the result together.
This is currently spreading as "this AI-generated pizza commercial" with no further explanation, but Tom's Hardware interviewed the actual people who made it work:
www.tomshardware.com/news/ai-p…
#PizzaAd #AIVideo #AI #PepperoniHugSpot #GPT #LLM #Runway #RunwayGen1 #RunwayGen2 #MidJourney #ElevenLabs #AIVoice #soundraw
So is did every online calculator built for SEO traffic dead now with GPT? We have been testing some retirement planning calculations out on GPT Chat and they look damn good even with sloppy prompting.
#OnlineCalculators #CalculatorApp #WebCalculators #SEOcalculator #MarketingCalculators #FinancialCalculators #ConversionCalculators #seo #gpt #gpt #financialwebsites
Could those with knowledge of LLMs share references/thoughts/sources about #WaterMarking (ie the embedding of a signal/pattern to flag non-human text/tokens)?
This - https://arxiv.org/pdf/2301.10226.pdf - is helping me understand but I have basic questions, eg.:
- *who* creates & embeds the watermark?
- Can a teacher do it?
- Would the watermark work on all LLMs (inc. commercial)?
I'm trying hard to understand 😣 😣 😣
@amills @emilymbender @timnitGebru #ArtificialIntelligence
#gpt #AcWri #AcademicWriting