#GPT4
Three Spanish MIT physics postdocs receive Botton Foundation fellowships https://ainewsfeed.io/article/0b7bb556ff1499ed63753ceea5aa013b #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Host ML models on Amazon SageMaker using Triton: ONNX Models https://ainewsfeed.io/article/d3cc47b15cbb0510dd51ca74d59c9826 #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Fast-track graph ML with GraphStorm: A new way to solve problems on enterprise-scale graphs https://ainewsfeed.io/article/cf133b3d91c90bda54b0c44b5a312f6d #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Eye in the Sky With AI: UCSB Initiative Aims to Pulverize Space Threats Using NVIDIA RTX https://ainewsfeed.io/article/5ac1101a0caf86889ee75e492c8e88aa #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2023 https://ainewsfeed.io/article/27d1f252dc9f35a95cf01f3d2963cb5b #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Vielleicht kann man #KI-Tools demnächst in Endlosschleifen rennen lassen…
(Wirklich neu ist das ja nicht, da #Langchain und Co. (inkl. #WolframAlpha-Plugin zu #GPT4) auch etwas Ähnliches leisten, es einfach deutlich weniger spektakulär vermarkten.)
https://www.heise.de/news/Googles-Bard-kann-besser-Rechnen-und-logisch-Antworten-9182607.html
🧠 Ho provato il plugin di #Noteable per #ChatGPT, e devo dire che è molto interessante.
💡 Come funziona? Una volta creato il progetto su Noteable e connessa l'origine dei dati è possibile interagire con i dati attraverso dei #prompt in linguaggio naturale da ChatGPT.
#AI #IntelligenzaArtificiale #GPT4 #LLM #DataAnalysis #dataviz #DataScience
Teaching robots to tidy up based on user preferences using large language models
#AI #ArtificialIntelligence #GPT4 #tech #technology #ChatGPT4 #news #ainews
https://search.ai.wiki/teaching-robots-to-tidy-up-based-on-user-preferences-using-large-language-models/
Get started with the open-source Amazon SageMaker Distribution https://ainewsfeed.io/article/4f440e4caeae01260c879069559f44ff #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
MIT researchers make language models scalable self-learners
#ChatGPT #OpenAI #AI #ArtificialIntelligence #GPT4 #tech #technology #ChatGPT4 #news #ainews
https://search.ai.wiki/mit-researchers-make-language-models-scalable-self-learners/
Eight (plus one) things everyone who uses #ChatGPT #GoogleBard #GPT4 #BingChat should know about #LargeLanguageModels
https://youtu.be/RX-gGs_EV7M
[#AI Coffee Break with Letitia]
New model offers a way to speed up drug discovery https://ainewsfeed.io/article/566ce597afcf11b446841c4aae6f5912 #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Bringing the social and ethical responsibilities of computing to the forefront https://ainewsfeed.io/article/24e212ab938f71d6279f01580385c70d #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Chatgpt designs a robot https://ainewsfeed.io/article/d4e6e827be8f8dc60cf795895f1f371a #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
A Gentle Introduction to Hallucinations in Large Language Models https://ainewsfeed.io/article/05037488948061bac6422ff82d4fbc9c #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
ChatGPT as Your Expert Helper https://ainewsfeed.io/article/20782c8421d1282420b56de87f23ca4f #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Exploring Generative AI in conversational experiences: An Introduction with Amazon Lex, Langchain, and SageMaker Jumpstart https://ainewsfeed.io/article/4905da3058f6c7fef68f13dec0435504 #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
MIT researchers make language models scalable self-learners https://ainewsfeed.io/article/c5186a2cede6d681019ac0f50f55422b #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Less Is More: A Unified Architecture for Device-Directed Speech Detection with Multiple Invocation Types https://ainewsfeed.io/article/511d22918ff39cf6e10bd566c2502faf #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Introducing popularity tuning for Similar-Items in Amazon Personalize https://ainewsfeed.io/article/823d5f5c25ed7f13ea2dd4f53be1d2fa #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Link-credible: Get in the Game Faster With Steam, Epic Games Store and Ubisoft Account Linking on GeForce NOW https://ainewsfeed.io/article/cfff76d8c7890790c08e73e1ff6e80b8 #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
A group of researchers have built an #AI bot that can play the popular #Minecraft game all by itself. Unlike other Minecraft agents that use reinforced learning techniques, the bot developed by a motley team uses #GPT4 language model.
https://trib.al/6lgQZxk #press
Accelerate PyTorch with DeepSpeed to train large language models with Intel Habana Gaudi-based DL1 EC2 instances https://ainewsfeed.io/article/b1799a687eb90a35aada204a90bf1fd0 #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
What Is Photogrammetry? https://ainewsfeed.io/article/aa1e6fea7f8eee121c92fdb7e1f965bb #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Taking AI to School: A Conversation With MIT’s Anant Agarwal https://ainewsfeed.io/article/b4fa80e5f70733272d7e3aa4502ea465 #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Prompt: fem diagrames amb un fantàstic connector de GPT-4
https://mestriacat.wordpress.com/2023/06/07/prompt-fem-diagrames-amb-un-fantastic-connector-de-gpt-4/
#iA #edtech #chatgpt #gpt4
Retrain ML models and automate batch predictions in Amazon SageMaker Canvas using updated datasets https://ainewsfeed.io/article/62df866bc0b766f5e81e08afa19bcde8 #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Expedite the Amazon Lex chatbot development lifecycle with Test Workbench https://ainewsfeed.io/article/4631b4848530fb3e7a4f586d2cc64bb4 #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Announcing enhanced table extractions with Amazon Textract https://ainewsfeed.io/article/2078b3c89e1d8b17e48ca3cf57b80d9a #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
NYU, NVIDIA Collaborate on Large Language Model to Predict Patient Readmission https://ainewsfeed.io/article/08b9330616de7d08e77378eb486eb373 #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Calling all #Minecraft fans! 🎉
Meet Voyager - an #LLM power agent that utilizes #GPT4 to engage in Minecraft gameplay. It demonstrates remarkable capabilities by learning, retaining knowledge & showcasing exceptional expertise in Minecraft.
Read more on #InfoQ and get ready for an unforgettable gaming experience: https://bit.ly/43JpgyP
Quantum computers are better at guessing, new study demonstrates https://ainewsfeed.io/article/63eede3e97f31dc21ad887155afe0186 #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Sponge makes robotic device a soft touch https://ainewsfeed.io/article/00347fa5f7f9082320e041de5772e824 #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Technology Innovation Institute trains the state-of-the-art Falcon LLM 40B foundation model on Amazon SageMaker https://ainewsfeed.io/article/15df445616cd08fac57904bc7b75d825 #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Q&A: Gabriela Sá Pessoa on Brazilian politics, human rights in the Amazon, and AI https://ainewsfeed.io/article/d3ca19fd3cf918a31f3670588311bc48 #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Build high-performance ML models using PyTorch 2.0 on AWS – Part 1 https://ainewsfeed.io/article/00759ad5673c80f66ca87398ad15eafb #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Efficient Multimodal Neural Networks for Trigger-less Voice Assistants https://ainewsfeed.io/article/364f618cd86894a3c79d3fd68e9f5c5c #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Flash Sale: Unlock Your AI Potential Today! https://ainewsfeed.io/article/ee47d3d08f5c3da5d3ec8747c883af40 #ai #tech #machinelearning #deeplearning #nlp #gpt4 #openai
Ëm mäin Job (zumindest en Deel dovunner) muss ech wuel awer fäerten. *
Extrem schnell en "Tutorial" geschriwwe fir en (einfachen) Bug, deen awer esou kloer net direkt iwwer Google ze fanne war.
Zesumme mam GitHub Copilot ass zumindest den Junior Dev domadder schon a Gefor 😩 *#GPT4
* Net seriö gemengt ;)
There is a lot going on with a large language models that most people are ignoring right now. For instance, the energy costs of running inference for millions of users.
This is WILD:
Carbon Health is using in-house #AI for medical notes
But it looks like they're just using #Amazon's Transcribe Medical speech rec tool, sending that transcript through #GPT4 (along with patient history) & calling it good?!
via my @STAT colleague Mohana Ravindranath
https://www.statnews.com/2023/06/05/ai-medical-records-carbon-health/
#healthtech #healthcare #health #medicine #artificialintelligence #doctors #hospitals
An AI team at Nvidia created a bot that uses GPT-4 to solve problems inside Minecraft.
https://www.wired.com/story/fast-forward-gpt-4-minecraft-chatgpt (4-min read)
#AI #ML #GPT4 #nvidia #tech #technology #minecraft #automation
ChatGPT kann jetzt auch Buchhaltung. Neue Version der KI besteht US-Examen zum Wirtschaftsprüfer und Steuerberater. #ChatGPT #KI #kuenstlicheIntelligenz #GPT4
https://www.scinexx.de/news/technik/chatgpt-kann-jetzt-auch-buchhaltung/
#AI #GenerativeAI #LLMs #Chatbots #GPT3 #GPT4 #OpenAI #Google: "The fundamental issue here is one of transparency. The builders of the big closed models—GPT-3, GPT-4, Google’s PaLM and PaLM 2, Anthropic’s Claude—refuse to tell us what’s in their training data.
Given this lack of transparency, there’s no way to confidently state that private data that is passed to them isn’t being used to further train future versions of these models.
I’ve spent a lot of time digging around in openly available training sets. I built an early tool for searching the training set for Stable Diffusion. I can tell you exactly what has gone in to the RedPajama training set that’s being used for an increasing number of recent openly licensed language models.
But for those closed models? Barring loose, high-level details that are revealed piecemeal in blog posts and papers, I have no idea what’s in them."
“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
🧠 Secondo uno studio di Rome Business School, il mercato italiano dell'#AI è cresciuto del +92% negli ultimi 3 anni.
🔗 La ricerca: https://romebusinessschool.com/blog/digitisation-big-data-and-ai-in-italy-digital-ethics-and-use-of-data/
Top 10 tools for detecting ChatGPT, GPT-4, Bard, and Claude. ~ Abid Ali Awan. https://www.kdnuggets.com/2023/05/top-10-tools-detecting-chatgpt-gpt4-bard-llms.html #ChatGPT #GPT4 #Bard #Claude
Der #Minecraft-Bot #Voyager arbeitet auf Basis von #GPT4 und verbessert mit der Zeit seinen eigenen Code. https://winfuture.de/news,136605.html?utm_source=Mastodon&utm_medium=ManualStatus&utm_campaign=SocialMedia
Is there research on how the consumption of generated content changes the perception or other skills/behaviors of individuals?
Is it more or less cognitively exhausting to consume generated content?
#Gpt #stablediffusion #midjourney #ChatGPT #LlMs #generativeart #Gpt4
That said, I've heard #GPT4 flexes pretty hard on #GPT3, so I might subscribe to ChatGPT proper for a month to get my measly 25 messages per 3 hours.
I've heard #Bing uses GPT4 so folks are using that as a sort of bypass, but I imagine there's a tradeoff they're making in i/o privacy and retention guarantees. Any other recommendations?
Welp, my daily driver search engine is #Google, and I've turned on #SGE (Search Generative Experience). It's fine, and sort of helpful, while feeling like they've bloated the tight search experience a bit with bolted-on SGE. (Not complaining, too much, it probably had to be done, and it's not bad, just not great.)
It's helpful, just not great. And #Bard? Really not as smart as #ChatGPT / #GPT4, and that really detracts.
#GPT4 niega los hechos:
De acuerdo con mis fuentes, el incidente ocurrió durante una simulación llevada a cabo por la Fuerza Aérea de los Estados Unidos. Un dron habilitado con IA “mató” a su operador humano para anular una posible orden de “no” que lo detuviera de completar su misión . Ninguna persona real resultó herida .
Wow, this #Voyager bot is incredible! It's amazing to see how much more efficient and productive it is compared to other AI bots. I'm sure this #GPT4 technology will revolutionize the way we do business in the future, by automating mundane tasks and freeing us up to focus on more complex, creative endeavors. #Nvidia #Minecraft #ChatGPT #AI http://www.techmeme.com/230602/p1#a230602p1
🧠 I ricercatori di Berkeley e Microsoft Research presentano #Gorilla: un modello di linguaggio che batte #GPT4 nella generazione delle chiamate API adeguate per rispondere alle richieste in linguaggio naturale.
🦾 Tecnicismi a parte, sistemi come questo, integrati su catene di pensiero, possono dare risultati straordinari.
🔗 Immaginiamo ad esempio integrato su #LangChain!
The post discusses the impact of GPT-4, the most advanced version of a transformer model, on the development of Generative AI tools. These tools create content mimicking a particular style using a self-attention mechanism. The post also highlights the potential fears associated with such tools. https://blog.cloudflare.com/secure-generative-ai-applications/ #GenerativeAI #GPT4 #TransformerModel #softcorpremium
The post discusses the rise of large language models like GPT-4 and their ability to generate fluent text, but also highlights their shortcomings in areas like chess and go, math, and coding. The author refutes claims of artificial general intelligence and notes the tendency to generate confidently incorrect text. The post ultimately... https://www.infoworld.com/article/3697272/which-is-better-generative-ai-or-reinforcement-learning.html#tk.rss_all #largeLanguageModels #GPT4 #AI #softcorpremium
The brexhq/prompt-engineering GitHub repository provides tips and tricks for working with large language models like OpenAI's GPT-4. The repository provides useful information for those working with GPT-4 or other large language models. https://github.com/brexhq/prompt-engineering #OpenAI #GPT4 #languageModels #softcorpremium
GPT-4 meets Minecraft! Researchers at Nvidia use the language model to create Voyager, a bot that solves problems in the game. It suggests goals, generates code, and learns from feedback. AI automation in gaming has huge potential. #GPT4 #Minecraft #AI #Automation #Nvidia
🧠 La supervisione del processo!
💡 Un nuovo studio di OpenAI che addestra un modello non solo con un feedback sui risultati, ma su tutti gli step della catena di pensiero.
🦾 Oltre a migliorare le performance, questo concetto va ad agire sull'allineamento e sul controllo, allontanando le allucinazioni (errori logici).
🔗 Il paper: https://openai.com/research/improving-mathematical-reasoning-with-process-supervision
#GPT4 hasn't degraded. It's just that people's initial impressions finally caught up with reality.
I'm very optimistic about LLMs. However, we should brace ourselves for another wave of startups that are shells for an online API, spending other people's money on building solutions that underdeliver and that nobody needs.
@randomgeek #Cody uses #Anthropic’s #Claude API or can be configured to prompt #OpenAI instead, as well as use the latter for embeddings: https://docs.sourcegraph.com/cody/faq#what-third-party-cloud-services-does-cody-depend-on-today
I’m glad neither #Sourcegraph nor its third-party #LLMs train on the code you give it: https://docs.sourcegraph.com/cody/faq#does-cody-train-on-my-code
Cody is probably prompting #GPT4 (OpenAI deprecated the API for the #LicenseWashing #Codex model behind #Copilot) based on this #Slack #bot code: https://sourcegraph.com/github.com/sourcegraph/sourcegraph@e772c7cbd177961cb2f8b02fcb35727158e8edef/-/blob/client/cody-slack/src/services/openai-completions-client.ts?L28
Large Language Models, Super Human Intelligence and the Methodological Hallucinations of Modern AI: https://open.substack.com/pub/artificialminds/p/large-language-models-super-human?utm_campaign=post&utm_medium=web
#ai #llms #existentialrisks #aiethics #largelanguagemodels #chatgpt #gpt4 #superintelligence #artificialintelligence
Today's phrase regarding #AI: Think
"Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war."
-- Sam Altman, CEO, OpenAI, et. al.
[Emphasis mine.] Keep this in mind when you think about the fun things you can do with the technology -- then think whether you should.
#boostingIsSharing #CommentingIsCool
#chatgpt #machinelearning
languagemodels #intelligence #largelanguagemodels #llms #llm #chatgpt #gpt4 #superintelligence #AIrisks #AIsafety
Come and apply your #GenerativeAI #MachineLearning skills to social good at Nesta! https://www.nesta.org.uk/procurement-and-invitations-tender/call-for-an-expert-to-harness-generative-ai-for-social-good/
#ML #AI #GPT #TechForGood #Jobs #LanguageModels #LargeLanguageModels #LLM #ChatGPT #Intelligence #ArtificialIntelligence #GPT4
Large Language Models, Super Human Intelligence and the Methodological Hallucinations of Modern #AI: https://lnkd.in/dbdkKD7z
#AI #languagemodels #intelligence #largelanguagemodels #llms #llm #chatgpt #gpt4 #superintelligence #AIrisks #AIsafety #cognitivedesignforartificialminds #esistentialrisks #computationalmodelsofcognition
ChatGPT per a iPhone, ja disponible a Espanya https://wwwhatsnew.com/2023/05/31/chatgpt-para-iphone-ya-disponible-en-espana/ #iA #chatgpt #gpt4
📊 Come possiamo analizzare i nostri dati con il supporto dell'#AI?
🧠 In questo post racconto come ho sfruttato #ChatGPT e #GPT4 per produrre il codice #Python per l'analisi di un dataset corposo.
🦾 Oltre ad estrarre dati, genero widget per una dashboard e clusterizzo le informazioni per avere il quadro completo.
#IntelligenzaArtificiale #DataAnalysis #DataScience #LLM #WebAnalytics #dataviz
https://www.alessiopomaro.it/chatgpt-python-analisi-dati-ai/
🧠 "Occorre consapevolezza dei cambiamenti, conoscenza concreta delle nuove tecnologie per una regolamentazione efficace e attenta al ruolo imprescindibile dell’essere umano in questo rapido processo di trasformazione economica e sociale".
🦾 Stasera, presso il Tecnopolo Tiburtino, si terrà "Sinergie Digitali: l’ascesa delle AI Generative", dove rifletteremo su questi temi.
🔗 Per chi non potrà essere presente: https://vimeo.com/event/3433440
@inthehands but the power of generative AI could also leverage people with smaller means than big corporate. I think it was too rapid the way #OpenAI rolls that out in public tho, and changed it as a for profit model. They should change their name. I prefer Open Source #LLM like #Vicuna that which is as good as #GPT3 and #GPT4.
https://lmsys.org/blog/2023-03-30-vicuna/
I don't think they will achieve #AGI tho, someday but it's too early #OpenAI are just an hype right now. It won't last.
Precious @emilymbender #scicomm skills in just 4 min:
1 #TextSynthesisMachine
2 #aiHype
3 #Anthropomorphization
4 #aiBias
5 #UndisclosedDatasets
fav bit:
» if it makes sense,
it is because you made sense of it «
HT via @WritingItReal
Hree are the presentation slides from Heiko Paulheim's keynote at the #text2kg workshop at #eswc2023 :
Knowledge Graph Generation from Wikipedia in the Age of ChatGPT: Knowledge Extraction or Knowledge Hallucination?
#elephantintheroom #knowledgegraphs #llms #gpt #gpt4 @eswc_conf
High nerdlevel but fun, self-exploring and developing GPT-4 based algorithm in a minecraft world.
#AI #GPT4 #Minecraft https://voyager.minedojo.org/
Automatic Aura
Jul 7 - Aug 27, 2023
Installation Space
FeministFuturist collective: Karen Meninno, Carolyn Wirth, Christina Balch, Marjorie Kaye, Homa Sarabi, Freedom Baird
Opening Reception: Friday, July 7, 4-8pm
49 Eagle Street, North Adams, MA 01247
#HarmonAI #GiftsForAIs #TheConcord #RobotChurch #WisdomNexus #Dodecahedron #AutomaticAura #Massachusetts #FeministFuturist #InstallationSpace #DoingWhatTheBotsSay #ChatGPT4 #ChatGPT #OpenAI #GPT4 #AI
Je moet niet teveel meegaan in de AI hype, natuurlijk is een tool/algoritme zoals ChatGPT niet "intelligent" in de menselijke zin. We worden echt niet allemaal vervangen, werk zal niet verdwijnen maar veranderen. Dat er wel echt iets fundamenteels aan de hand is illustreert het vermogen van ChatGPT (GPT4 / betaalde versie), om dingen te doen die tot voor kort onmogelijk werden gehouden https://neerlandistiek.nl/2023/05/nieuwste-chatbot-haalt-het-vwo-examen-wel-met-vlag-en-wimpel-wat-nu/
George Orwell recognised back in 1945 that a mechanization process in visual and written arts was already underway. He was convinced that, under corporate bureaucracy and/or the totalitarian state, mechanized writing would eliminate imagination from the creative process. The product, Orwell predicted, would only be rubbish. But that might not matter... #writersstrike
#Artificialintelligence #LargeLanguageModels #GPT4 #Orwell #garbageingarbageout https://blackd.substack.com/p/can-machines-write-books-that-arent
Meta AI Unleashes Megabyte, a Revolutionary Scalable Model Architecture - Artisana
"A #Meta team of #AI researchers has proposed an innovative architecture for AI models, capable of generating expansive content in text, image, and audio formats, stretching to over 1 million tokens. This groundbreaking proposal, if embraced, could pave the way for the next generation of proficient AI models, transcending the Transformer architecture that underpins models such as #GPT4 and #Bard, and unleashing novel capacities in content generation."
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
LogiCoT: Logical Chain-of-Thought instruction-tuning data collection with GPT-4. ~ Hanmeng Liu, Zhiyang Teng, Leyang Cui, Chaoli Zhang, Qiji Zhou, Yue Zhang. https://arxiv.org/abs/2305.12147 #GPT4 #Logic
Experimental results from applying GPT-4 to an unpublished formal language. ~ Gregor vom Scheidt (@GregorVScheidt). https://arxiv.org/abs/2305.12196 #ChatGPT #GPT4 #FunctionalProgramming #Logic
#AI #GenerativeAI #LLMs #Chatbots #ChatGPT #GPT4: "It sounds like you don’t think GPT-5 or GPT-6 is going to make a lot of progress on these issues.
Brooks: No, because it doesn’t have any underlying model of the world. It doesn’t have any connection to the world. It is correlation between language.
By the way, I recommend a long blog post by Stephen Wolfram. He’s also turned it into a book.
I’ve read it. It’s superb.
Brooks: It gives a really good technical understanding. 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."
🎯 Explainable AI suffers from an epidemic. I call it Explainability Washing.
💡Think of it as window dressing—techniques, tools, or processes created to provide the illusion of explainability but not delivering it
Let’s use this hyped example from OpenAI. Title is sensational-- Language models can explain neurons in language models.
But is that the case? Let's dig in. 👇
https://openai.com/research/language-models-can-explain-neurons-in-language-models
#AI #ML #XAI #gpt4 #chatGPT #Bard #LLM #NLP #HCI #UX #UI #responsibleai #mastodon
1/n
Quick #AI tip. I recorded a podcast and created a transcript of the content. I then had #ChatGPT with #GPT4 turn that transcript in to a blog post. Interestingly, it formatted the text with headings and other HTML elements. I was able to quickly copy the text and paste it in to WordPress, and the text was properly formatted so this means that ChatGPT is now formatting text so that it can be used with a blogging application. That is amazing!
Ok, my latest project:
Send voice messages to #gpt4 via a #free #telegram bot while driving, jogging, cleaning etc:
https://t.me/magicmemosbot
(DM for an invite)
I was trying to demonstrate how ChatGPT can emulate a given style and used this example. #AI #GPT4 #Bogart #SamSpade #SanFrancisco
So I had #ChatGPT and #GoogleBard both write a #Destiny2 ending story for Zavala to compare which #AI could write the best story. ChatGPT with #GPT4 wrote a fairly good story, and Bard tried, but it didn’t write a story based on the prompt. So, I have to give the win to ChatGPT.
The satirical news site Waterford Whispers News pokes fun at The Irish Times naming the persona of the hoax opinion piece; Adriana Acosta-Cortez, as their new Opinion editor.
#Ireland #WaterfordWhispersNews #WWN #TheIrishTimes #journalism #AI #ChatGPT #GPT4
Ireland's newspaper of record, The Irish Times got duped by an opinion piece written mostly by generative AI; GPT-4. The piece was removed by the paper and an apology posted by the editor.
Original opinion piece: https://web.archive.org/web/20230511121811/https://www.irishtimes.com/opinion/2023/05/11/irish-womens-obsession-with-fake-tan-is-problematic/
Editorial apology: https://www.irishtimes.com/ireland/2023/05/14/a-message-from-the-editor/
More analysis: https://www.theguardian.com/media/2023/may/14/irish-times-apologises-for-hoax-ai-article-about-womens-use-of-fake-tan
#Ireland #TheIrishTimes #journalism #ai #GenerativeAI #ChatGPT #GPT4
「 Any C-Suite executive who thinks they can replace software engineers (even novices) with generative AI will be at a disadvantage compared to competitors who use it to empower software engineers 」
—neverworkintheory.org
#GenerativeAI #Automation #GPT4 #ChatGPT
https://neverworkintheory.org/2023/05/13/generative-ai.html
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
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/
Here we jump to the end of the discussion, in which GPT-4 pretty much laid out a plan for humans to take advantage of our natural advantages and reverse some of our own most dangerous programming. #AI #GPT4 #Humanity #Nature
https://www.seldoncrisis.net/ai-and-humanity-a-conversation-with-gpt-4/
Episode 40 of #SeldonCrisis features a conversation with GPT-4 on a fairly complex topic, the irreplaceability of human beings. This clip sets the stage, and a couple of others give a little more taste for the conversation itself. #AI #GPT4 #LLM
https://www.withfanfare.com/p/seldon-crisis/ai-and-humanity-a-conversation-with-gpt-4
Biden ft. Trump - Ni**as In Paris (AI Cover) https://youtu.be/sKqfeWQOOZk
#ai #aiart #aiartwork #chatGPT #gpt4 #lmm #generativeai #generativeart #music #biden #trump
Latest #SeldonCrisis episode is a conversation with GPT-4 on what makes humans unique and irreplaceable... or what doesn't? #AI #GPT4 #LLM
https://www.withfanfare.com/p/seldon-crisis/ai-and-humanity-a-conversation-with-gpt-4
🚀 I'm excited to share my latest blog post, "Introducing Rbot: A Personalized AI Assistant, Written by Rbot"! Check it out here: https://rajiv.com/blog/2023/05/08/introducing-rbot-a-personalized-ai-assistant-written-by-rbot/
In this post, I detail my journey in creating Rbot, an AI-powered chatbot that serves as a personalized assistant. I've built Rbot using the GPT-4 model provided by OpenAI. It's designed to be an open-source project for others to adapt and create their own intelligent assistants. 🤖
#technofeudalism #chatgpt #ai “#OpenAI, a company that is neither open nor actually about any intelligence, made its entire business model about violating as many licenses as possible … Now they are threatening an #opensource developer for creating a tool that bypasses their fees by combining results from different, free APIs which use #GPT4 underneath.” https://d-s.sh/2023/today-in-ai-open-ai-is-driking-its-own-poison/
🤖 rbot: Rajiv's chatbot utilizing the GPT-4 model to offer engaging conversations with a personalized touch and advanced context understanding.
🌐 Explore the project on GitHub and share your thoughts: https://github.com/rajivpant/rbot
Listening to very smart people talk about #GPT4 I'm reminded of the joke about a checkers-playing dog.
A guy has a dog that plays checkers. "My goodness," everyone says, "that's amazing. What a brilliant dog!"
"Not really," he replies, "I beat him four games out of five."
That's GPT4. It's capacities are amazing and completely unexpected.
But it's also so limited. You shouldn't back the dog in a checkers tournament, and you shouldn't use an LLM as a medical assistant or in many other ways.
After hearing Sebastian Bubeck talk about the #SparksOfAGI paper today, I decided to give #GPT4 another chance.
If it can really reason, it should be able to solve very simple logic puzzles. So I made one up. Sebastian stressed the importance of asking the question right, so I stressed that this is a logic puzzle and didn't add anything confusing about knights and knaves.
Still, it gets the solution wrong.
HuggingChat, an open-source ChatGPT alternative:
#GPT4 does not give any more information about me anymore. I guess they are being cautious about personal information now.
Using #GPT4 to code in a new language is cool, but have you used it to resume a once-promising relic that 2was left to wither in chaos for half a decade?
I can't be the only one who feels the need to thank #GPT4 every time a 15 minute sesh saves me half a day of coding.
Imagine you're blind, and you're excited because this very thought-provoking article you're reading is about to break down a bunch of covid data to you in a chart. You then get down to the alt text for said article, and all the alt text says is 'Chart showing the numbers of covid cases in each state from the years 2020-2023...' The article then presumes to assume that you can read all the data points on the chart. You'd be thinking, 'Well thank you for providing alt text, but I still have no data points.' You'd be frustrated because the article is referencing data points to which you have no access. You might even think, 'Well, I appreciate the conclusions brought out in this article, but without accessing the chart, I can't fact check the statistics mentioned by the article's author.'
Moral of the story? Charts absolutely deserve good alt text.
Pascal Laliberté, I'd like to have text descriptions above charts. Before today, generated descriptions were tough.
Charts deserve good alt text!
Soon I think this will be everywhere with the help of LLMs. For now, it's slow-ish, costly-ish, but possible with OpenAI's gpt-4 model.
My little experiment got me this description for the chart below:
"Team met 4.5-hour first response goal on average. Maximums occasionally exceeded goal, with highest at 6.57 hours."
Not bad
#llms #ai #openai #gpt4 #accessibility Media description: Screenshot of a chart showing the time to first response to support cases, the min/max range and average daily, plus a chart for support case volume below.
3 hours ago. Public.