Both free and paid ChatGPT users can now access ChatGPT-4o Mini instead of GPT-3.5, released in November 2022. GPT-4o Mini supports text and vision via the OpenAI API, enabling developers to create new applications using OpenAI technology.
MULTIMODAL AI
-
GPT-4o Mini Multimodal Capabilities and Features
By
–
4/ Multimodal Magic: Text? Check. Vision? Check. Audio and video support coming soon! GPT-4o Mini is your all-in-one AI powerhouse.
-
JASCO: Music Generation Model with Improved Controllability
By
–
We recently announced JASCO, a music-generation model with improved controllability using conditioning inputs like chords or beat. We've shared more details on this work in the research paper https://
go.fb.me/sou2ox -
Hyper-personalization at Scale: AI Speed Breakthroughs
By
–
Not sure you had a chance to read the article, @WarrenWhitlock
, but the author agrees with you. It's about hyper-personalization at scale and at speeds we haven't seen before. -
Real AGI: Building Superhuman Brains for Any System
By
–
Real AGI is knowing how to build a superhuman brain for any set of sensors and actuators.
-
GPT-4o Voice Mode Alpha Launch This Month
By
–
La alpha del modo voz de GPT-4o empieza este mes. Quiénes accederán y cuántas "semanas" hasta tener acceso general son incógnitas por resolver.
-
How to Identify AI-Generated Music: Detection Methods
By
–
What makes you think/how do you know this is AI-generated? This track (and some of their others) sound like people sang them — maybe there’s something I’m not catching on to?
-
GPT-4o mini now available on Poe platform
By
–
GPT-4o mini is available on Poe! GPT-4o mini is OpenAI’s newest model and the highest performing small model available. Benchmark evaluations show it surpasses other compact models in reasoning (text and vision), math, and coding tasks. (1/2)
-
Meta Joins Tech Companies Restricting AI Release in Europe
By
–
Europe’s AI suicide gathers steam: Meta is the latest tech company to say it won’t release its AI there. https://
techopedia.com/news/meta-says
-it-wont-release-future-multimodal-ai-models-in-eu
… -
DALL•E 3 Quietly Improves Text Generation Capabilities
By
–
Oye… DALL•E 3 como que con el tiempo ha ido mejorando de forma silenciosa? Me ha sacado el texto a la primera. Y por lo que confirman no es el modelo multimodal de GPT-4o (que sí sabemos podrá escribir) sino la misma arquitectura de DALL•E de siempre. Probad a ver