- AI Valley
- Posts
- 🐶 AI could make KYC useless
Together with
Howdy, AI Family!
Welcome to another edition of AI Valley. We're thrilled to share some fascinating AI stories and updates with you.
Here’s what’s on the menu today:
⚖️ OpenAI’s response to NYT lawsuit
🤖 AI could make KYC useless
🤩 No more waitlist for Pika
🌐 Tools, Tweets, and Resources
Ready, Set, Go: Today’s read should be just about a 3.5-minute read.
UPDATE
OpenAI responds to The New York Times' lawsuit, emphasizing its commitment to ethical AI and partnerships with news companies.
The firm argues that using public content for AI training is fair use and offers an opt-out for publishers.
OpenAI suggests The New York Times might have misused the AI to produce repeated content, but remains open to collaboration despite the legal challenge.
TOGETHER WITH OCTAML
Fine-tuned LLMs are the best way to create highly customizable applications, and a smaller fine-tuned Mixtral 7B can even outperform larger models like GPT 4 in both quality and cost.
But popular LLM serving platforms charge 2x for inference APIs on fine-tuned models. Say no to costly service and bring your fine tunes to OctoAI!
Host any fine-tuned Llama 2, Code Llama, or Mi(s/x)tral models on OctoAI
Run inferences at the same low per-token pricing as the default Chat and Instruct models in OctoAI
Experience the same super-fast latency available for default models
Read the blog to learn about more or sign up and try the default LLMs for free.
UPDATE
AI poses new risks. Attackers can use AI tools to modify a selfie and create a fake ID image, potentially passing KYC checks.
Tutorials online show how Stable Diffusion can be used to create artificial images of people in any setting, like a living room.
And now an attacker can adjust the images until it looks like the person is holding an ID. Then, they can use photo editing software to add a real or fake ID into the hands of the deepfaked person.
AI will rapidly accelerate broad use of private key cryptography and decentralized ID.
Check out this Reddit "verification post" and ID made with Stable Diffusion. When we can no longer trust our eyes to ascertain whether content is genuine we'll rely on applied cryptography.
— Justin Leroux (@0xMidnight)
5:52 PM • Jan 5, 2024
UPDATE
Pika recently dropped the waitlist.
You can go to Pika.art now and start creating videos for free!
TRENDING AI TOOLS
Blocks - An AI-powered email builder that will help you make stunning emails in minutes!
Stylar - A new Tool for AI Artist.
Listnr -A free voice & video generator with over 900+ voices in 142 languages.
Zocket - A generative AI model trained with local and global purchase data for the highest level of marketing performance.
Jan - An open-source chatbot that runs 100% offline on your computer and provides 10x productivity.
OpenVoice - A tool to clone voices with style control and multilingual support.
AI DEMO
DragNUWA is a video generation model that utilizes text, images, and trajectory as three essential control factors to facilitate highly controllable video generation from semantic, spatial, and temporal aspects.
HOT OFF THE PRESS
OpenAI is set to launch its GPT Store next week.
MyShell just launched a tool called OpenVoice, allegedly able to clone voices based on very short input clips.
Volkswagen is bringing ChatGPT into its cars.
Microsoft is integrating AI into computers by announcing a new Copilot key for keyboards.
LG announces its Al Smart Home Agent at CES2024
Amazon turns to AI to help customers find clothes that fit when shopping online
TWEET POST OF THE DAY
This video has also been translated into other languages so naturally it is hard to tell which is the original image.
THAT’S ALL FOR TODAY
Like what you see? Subscribe Now or Partner With Us
If you have anything interesting to share, please reach out to us by sending us a DM on Twitter: @heyBarsee
What did you think of today's newsletter: Your feedback helps to create better content for you. |
Together with
Howdy, AI Family!
Welcome to another edition of AI Valley. We're thrilled to share some fascinating AI stories and updates with you.
Here’s what’s on the menu today:
⚖️ OpenAI’s response to NYT lawsuit
🤖 AI could make KYC useless
🤩 No more waitlist for Pika
🌐 Tools, Tweets, and Resources
Ready, Set, Go: Today’s read should be just about a 3.5-minute read.
UPDATE
OpenAI responds to The New York Times' lawsuit, emphasizing its commitment to ethical AI and partnerships with news companies.
The firm argues that using public content for AI training is fair use and offers an opt-out for publishers.
OpenAI suggests The New York Times might have misused the AI to produce repeated content, but remains open to collaboration despite the legal challenge.
TOGETHER WITH OCTAML
Fine-tuned LLMs are the best way to create highly customizable applications, and a smaller fine-tuned Mixtral 7B can even outperform larger models like GPT 4 in both quality and cost.
But popular LLM serving platforms charge 2x for inference APIs on fine-tuned models. Say no to costly service and bring your fine tunes to OctoAI!
Host any fine-tuned Llama 2, Code Llama, or Mi(s/x)tral models on OctoAI
Run inferences at the same low per-token pricing as the default Chat and Instruct models in OctoAI
Experience the same super-fast latency available for default models
Read the blog to learn about more or sign up and try the default LLMs for free.
UPDATE
AI poses new risks. Attackers can use AI tools to modify a selfie and create a fake ID image, potentially passing KYC checks.
Tutorials online show how Stable Diffusion can be used to create artificial images of people in any setting, like a living room.
And now an attacker can adjust the images until it looks like the person is holding an ID. Then, they can use photo editing software to add a real or fake ID into the hands of the deepfaked person.
AI will rapidly accelerate broad use of private key cryptography and decentralized ID.
Check out this Reddit "verification post" and ID made with Stable Diffusion. When we can no longer trust our eyes to ascertain whether content is genuine we'll rely on applied cryptography.
— Justin Leroux (@0xMidnight)
5:52 PM • Jan 5, 2024
UPDATE
Pika recently dropped the waitlist.
You can go to Pika.art now and start creating videos for free!
TRENDING AI TOOLS
Blocks - An AI-powered email builder that will help you make stunning emails in minutes!
Stylar - A new Tool for AI Artist.
Listnr -A free voice & video generator with over 900+ voices in 142 languages.
Zocket - A generative AI model trained with local and global purchase data for the highest level of marketing performance.
Jan - An open-source chatbot that runs 100% offline on your computer and provides 10x productivity.
OpenVoice - A tool to clone voices with style control and multilingual support.
AI DEMO
DragNUWA is a video generation model that utilizes text, images, and trajectory as three essential control factors to facilitate highly controllable video generation from semantic, spatial, and temporal aspects.
HOT OFF THE PRESS
OpenAI is set to launch its GPT Store next week.
MyShell just launched a tool called OpenVoice, allegedly able to clone voices based on very short input clips.
Volkswagen is bringing ChatGPT into its cars.
Microsoft is integrating AI into computers by announcing a new Copilot key for keyboards.
LG announces its Al Smart Home Agent at CES2024
Amazon turns to AI to help customers find clothes that fit when shopping online
TWEET POST OF THE DAY
This video has also been translated into other languages so naturally it is hard to tell which is the original image.
THAT’S ALL FOR TODAY
Like what you see? Subscribe Now or Partner With Us
If you have anything interesting to share, please reach out to us by sending us a DM on Twitter: @heyBarsee
What did you think of today's newsletter: Your feedback helps to create better content for you. |