It "feels" GPT-3.5+ class, like Claude 2.1 like so far (though its reported benchmarks are closer to Claude 3/GPT-4 class). But that is just first impressions.
GENERATIVE AI
-
Surprise LLM Announcement Catches AI Community Off Guard
By
–
Did anyone expect a big new LLM coming? I was surprised by this announcement.
-
New Multimodal Model Competes with Gemini Video Capabilities
By
–
Interesting to see a big new multimodal model come out, especially as Gemini 1.5 has been the only publicly available model that could work with video
— Ethan Mollick (@emollick) 15 avril 2024
I have only played with it a bit, and the model seems pretty capable, but it hallucinates much more than Gemini on video so far https://t.co/WiI3JZUJhM pic.twitter.com/VLorP9KzOrInteresting to see a big new multimodal model come out, especially as Gemini 1.5 has been the only publicly available model that could work with video I have only played with it a bit, and the model seems pretty capable, but it hallucinates much more than Gemini on video so far
-
OpenAI Launches Batch API with 50% Cost Savings
By
–
Introducing the Batch API: save costs and get higher rate limits on async tasks (such as summarization, translation, and image classification). Just upload a file of bulk requests, receive results within 24 hours, and get 50% off API prices: https://
platform.openai.com/docs/api-refer
ence/batch
… -
OpenAI Announces Batch Product for LLM Usage
By
–
Lol last week's issue was about how most LLM usage is batch right now and openai announced a batch product today
-
BigTech AI Content Moderation: Control Consolidation Without Accountability
By
–
TLDR: to 'solve' trash content on trash surveillance platforms, BigTech AI will flag whether content is 'good' or not. Can we please stop falling for this? This isn't 'solving' the problem. It's simply consolidating control over platforms w/o attending to who will wield it.
-
AI Security Risk: Prompt Injection in Context-Aware Systems
By
–
"I have access to your email, I have access to your Slack messages — when you get a message for which the context of your past can help answer it, I can just give you that draft." Hope they're thinking hard about prompt injection!
-
ChatGPT Code Interpreter Alternatives With Internet Access
By
–
What are the best current alternatives to ChatGPT Code Interpreter? I want almost the exact same product but with the ability to access the internet to install additional packages or interact with APIs I'd prefer hosted, but still interested in hearing about good local options
-
Large-scale data processing for generative AI model training
By
–
Yeah, anything that has to do with processing large amounts of data. I actually think this would be amazing for helping train new models as well. You could use large batches of image classification and then use the results to train a model at the quality level of DALL-E.
-
OpenAI Batch API: 50% discount with 24-hour processing
By
–
50% discount on OpenAI pricing if you submit a batch and give them up to 24 hours to run the prompts for you – against any of their /v1/chat/completions models The batch format is newline-delimited JSON like this: https://
platform.openai.com/docs/api-refer
ence/batch/requestInput
…