Our OpenAI DevDay Breakout Sessions are now live on YouTube. Check out deeper dives of our new features for developers, techniques for maximizing LLM performance, and more.
GENERATIVE AI
-
Text-to-Speech Limitations and Unexpected Behavior Exploration
By
–
I’ve thrown a lot of weird prompts at text-to-speech and as yet haven’t got it to do anything unexpected. You can get it to [beep] its own swears, and occasionally an interjection might half work. Not as wild as something like Bark though.
-
Whisper hallucination demo: ChatGPT-4 chat with baby wanting more milk
By
–
Demo of Whisper hallucination — chat between ChatGPT 4 and a baby who wants more milk even though she just had two bottles:
-
Self-Destructing Models: Preventing AI Jailbreaking and Misuse
By
–
If ChatGPT can be jailbroken with only 20 cents, what can creators do to prevent their models from being repurposed for harm? Researchers describe one initial direction for increasing the costs of doing so: the self-destructing model.
-
Self-Attention Mechanism Implementation for Language Models
By
–
Understanding and Coding the Self-Attention Mechanism of Large Language Models From Scratch https://
bit.ly/3YL821l #AI #DeepLearning #MachineLearning #DataScience -
Google Researchers Present Latest Advances in ML and AI
By
–
Yesterday’s Research@ Seoul event was a success — Google Researchers led lightning talks on ML fairness, Genomics, Materials Science, Generative AI in Search, & Quantum. We're grateful to them for presenting their work & fueling progress in the field!
-
LangChain and Microsoft Azure announce strategic collaboration
By
–
LangChain Microsoft Today, we’re thrilled to announce our collaboration with @microsoft @azure
. Our joint customers will enjoy deeper product integrations that live up to our commitment to AI with the enterprise assurances the Azure ecosystem provides. more on the blog and -
GPT-4 Turbo Performance Doubles: Major Impact on User Experience
By
–
“From my own runs using our internal data, GPT-4 Turbo is more than twice as fast as GPT-4. In terms of product impact, this is big — big-big. […] The latency to first word or output seen by the user is drastically decreased with Turbo.” — @michaelvillar https://
michaelvillar.substack.com/p/gpt-4-turbo-
is-changing-the-game
… -
Keller Williams accelerates AI use cases with DataRobot platform
By
–
“We use DataRobot because it just works. We can rapidly build new AI use cases and have the optionality we need, especially with generative AI, that otherwise would not be possible.” Fred De Letter, Senior Director of Business Insights and Analytics, Keller Williams Realty, Inc.… pic.twitter.com/iHxWOxp2wn
— DataRobot (@DataRobot) 15 novembre 2023“We use DataRobot because it just works. We can rapidly build new AI use cases and have the optionality we need, especially with generative AI, that otherwise would not be possible.” Fred De Letter, Senior Director of Business Insights and Analytics, Keller Williams Realty, Inc.
-
Amdocs Builds Custom LLMs for Telecom Generative AI
By
–
.
@Amdocs to build custom #LLMs to allow communications service providers to efficiently deploy #generativeAI use cases across their businesses, from customer experiences to network operations and provisioning, using NVIDIA AI foundry service on @Azure
. https://
nvda.ws/3MOqSBh