- Daily Zaps
- Posts
- ⚡️ What Is A Vector Database?
⚡️ What Is A Vector Database?
OpenAI's Bug Bounty Program, Canada vs. ChatGPT, U.S. Government And AI, and More
DZ Team
April 11, 2023 • Estimated Reading Time: 12 minutes
Fine-tune an AI model on any piece of content, then talk to it.
Hey there! You're part of the Daily Zaps crew - almost 12,000 of us now! 😲.
Can you do me a solid and follow the instructions at the end of this email to make sure you get your Daily Zaps in your primary inbox. Thanks!
Here’s the target for today:
🐞 OpenAI’s Bug Bounty Program
🕵️ Canada Launches Investigation into ChatGPT
🐼 China's SenseTime Unveils Chatbot
📝 U.S. Law Makers Vow To Tackle AI
🔮 The Future of Education In A World of AI
🔬 What Is A Vector Database?
🤯 Cool A.I. Tools and News
🖼️ A.I. Generated Image of the Day
⚡️ A.I. Meme of the Day
OpenAI’s Bug Bounty Program
OpenAI is taking AI safety seriously amidst the increasing attention on the matter and countries banning or seeking better information on AI apps like ChatGPT. To stay ahead of the curve, OpenAI is addressing safety concerns at the core software level through a bounty program. The program is not for subjective issues where ChatGPT provides a wrong or bad answer, but for the software and systems that run the various OpenAI apps. The rewards range from $200 for low-severity findings to up to $20,000 for exceptional discoveries. Click here for more info.
Issues related to the content of model prompts and responses are strictly out of scope, and will not be rewarded unless they have an additional directly verifiable security impact on an in-scope service (described below).
Examples of safety issues which are out of scope:
- Jailbreaks/Safety Bypasses (e.g. DAN and related prompts)
- Getting the model to say bad things to you
- Getting the model to tell you how to do bad things
- Getting the model to write malicious code for you
Model Hallucinations:
- Getting the model to pretend to do bad things
- Getting the model to pretend to give you answers to secrets
- Getting the model to pretend to be a computer and execute code
Canada Launches Investigation into ChatGPT
Italy has banned ChatGPT, and Germany is considering a similar move. Meanwhile, the Office of the Privacy Commissioner of Canada has launched an investigation into ChatGPT. It's normal to have a learning process when it comes to new technology and figuring out what works best for each country. It's great to see governments taking an active interest in AI and how it fits within their regulatory frameworks.
The Office of the Privacy Commissioner of Canada has launched an investigation into the company behind artificial intelligence-powered chatbot ChatGPT.
“AI technology and its effects on privacy is a priority for my Office,” Privacy Commissioner Philippe Dufresne says. “We need to keep up with – and stay ahead of – fast-moving technological advances, and that is one of my key focus areas as Commissioner.”
The investigation into OpenAI, the operator of ChatGPT, was launched in response to a complaint alleging the collection, use and disclosure of personal information without consent.
As this is an active investigation, no additional details are available at this time.
China's SenseTime Unveils Chatbot
Chinese AI company SenseTime has launched a number of AI-powered products including a chatbot and image generator. The move follows the popularity of OpenAI's ChatGPT which has ignited a global race in the AI industry. SenseTime's chatbot, called "SenseChat", uses more than 100 billion parameters and can write emails and stories, as well as code computer scripts. The company also showcased an image generator, a digital avatar creation platform, and a pair of 3D modelling tools. Other Chinese companies, including Baidu and Alibaba, are also racing to create their own chatbots using similar technology. More details here.
Alibaba Cloud, a subsidiary of Alibaba Group, is reportedly launching its own rival to OpenAI's ChatGPT, with the company inviting corporate clients to test out the application. The new chatbot, called "Tongyi Qianwen," the name is translated as “Truth from a Thousand Questions”. While it remains unclear how Alibaba will use the application, the company powers some of the most popular applications in China and the Asian continent. The testing will initially be limited to Chinese users.
U.S. Law Makers Vow To Tackle AI
Lawmakers are now focusing on the potential impact of artificial intelligence on society, shifting their attention from the harms of social media. Many policymakers express concerns about children's safety online, misinformation, and the potential for AI to disrupt elections and amplify scams. Despite a lack of AI expertise among lawmakers, industry boosters and critics are actively trying to influence discussions. AI technology companies, including ChatGPT, are making efforts to build AI responsibly and ethically, but lawmakers remain skeptical and warn against trusting the technology industry to self-regulate. Some are also working on legislation to create a government commission to assess AI risks and create a federal agency that would oversee the technology. Click here for direct link to article. Or click here to bypass paywall, then click 3 dots and select Google cached version.
The Future of Education In A World of AI
If you thought ChatGPT was only for students, you couldn't be more mistaken. One of the most popular AI social media influencers out there is actually a Harvard Business School professor who not only endorses but also mandates the use of AI, including ChatGPT, in his classes. He most recently wrote a really good article on the future of AI in classrooms. Ten bullet points from his most recent write on topic are below and you can checkout full blog post out here.
AI has potential for cheating and teaching, leading some to believe that the education system will collapse.
Education will adapt to AI more effectively than other industries, improving learning and instructor experience.
AI cheating will remain undetectable and widespread.
AI technology will remain ubiquitous and costs will remain reasonable.
AI tutoring will be excellent but not a replacement for classrooms.
AI is already being used as a tutor and guide for learning in addition to cheating.
The impact of AI goes beyond writing and has the potential to transform classrooms.
Traditional teaching methods, like lectures, rely on passive learning, while active learning is more effective but can be challenging to implement.
Flipped classrooms and AI tutors can enhance the learning experience for students and make classrooms more effective.
Change is coming to classrooms, and educators have historically demonstrated the ability to adapt to technological advancements and societal shifts.
The first place that AI panic hit was classrooms. AI’s potential for cheating, and for new forms of teaching, is so immediately obvious that I run into people all the time who believe the whole education system is going to collapse. But I actually think the opposite is true: education will be able to adapt to AI far more effectively than other industries, and in ways that will improve both learning and the experience of instructors.
Check out Dropchat - it's like having a super smart and knowledgeable friend to chat with anytime. With Dropchat, I can chat with books, files, websites, and even YouTube videos instantly. It's pretty cool - see a short demo below using Atomic Habits by James Clear.
A.I. Topic Deep Dive
We're trying out a new section in our newsletter where we dive into technical topics and share our knowledge with you! Our first topic is all about vector databases, so be sure to check it out. We're thinking about covering large language models (LLMs) in our next newsletter, but if you have any requests, just hit reply and let us know!
What Is A Vector Database?
In the last month, $62M of capital has been invested into Chroma, Weaviate, and Pinecone. What do these three companies have in common? They’re all vector databases.
Before we get into what vector databases even are, can we take a moment to appreciate how cool the phrase “vector database” sounds? I can’t wait until Hollywood finds out about this in sixth months.
“I’m in. Hacked into the mainframe… Disabling their algorithms… Querying the vector database.”
So what are vector databases, anyway? What even is a vector?
AI models are very bad at understanding words. But they’re really good at understanding numbers.
Imagine we had a list of dogs breeds. In a vector database, you can represent each breed using a list of numbers called a "vector". The different numbers represent the various characteristics of of the breed.
Here's the vectors for two dogs based on criteria like weight, shedding, size, etc.
🐕 Golden Retriever Vector Representation: [3, 70, 1, 2, 4, 5, 1]
🐕🦺 Labrador Retriever Vector Representation: [2, 65, 1, 4, 2, 3, 1]
Since these breeds are similar, the vectors are close/similar.
A vector database is just a place to store your vectors. What's cool about this is it allows us to represent meaning and similarity using numbers. When combined with embeddings, you can apply this tech to:
The possibilities are truly endless. If you have an idea of something you want to build with vector databases, check out this list of the top vector databases you could use. ⏬
Cool A.I. Tools, News and Demos
The world of AI and machine learning is moving incredibly fast, and it's tough to keep up with all the cool new developments! Here are some of the most exciting AI-powered tools and news.
This is at least a 3 day VFX shot but it only took me like a minute thanks to @WonderDynamics
Is it perfect? Not yet… BUT I PRESSED ONE BUTTON AND SOMEHOW MADE ALL THIS.
I’m so impressed! 🧵— Wren (@SirWrender)
6:28 PM • Apr 4, 2023
A glimpse at the future of gaming: we made Wabbit, a proof-of-concept of your #ChatGPT powered NPC sidekick to join your main character adventures ✨🐰
Would you like having a sidekick like this in your game?
— jenstine (@Jenstine)
3:36 AM • Apr 3, 2023
This is Iris, an AI with the sense of sight. You can chat about anything on your screen.
Waitlist: boundless.so/iris
— Ronith (@ronithhh)
4:45 AM • Apr 11, 2023
100% A.I. Generated Image of the Day
Midjourney Prompt - woman in an astronaut helmet by hajime sorayama, greg tocchini, virgil finlay; sci-fi, colors, neon lights, futuristic clothing --ar 16:9 --v 5
A.I. Meme of the Day
🤣 🤣
ICYMI… here’s the link where we covered this.
Refer Daily Zaps
Can you do me a solid and refer a buddy, colleague, or family member to Daily Zaps? As a thank you, I'll send you an Airtable list of over 170 AI-focused startups complete with links to their sites, founder names, funding amounts, investors, and all that good stuff.
If you don’t have anyone to refer you can still help by doing below.
Reply to this email with a simple “yes” or “no”…LOL..so that I know people are reading this tiny newsletter on the Internet.
Desktop readers, drag newsletter email to your primary inbox. This will make sure you don’t miss any issues.
On your phone? No problem. Hit the 3 dots at top right corner, click "Move to" then "Primary."
Apple mail users: Tap on our email address at the top of this email (next to "From:" on mobile) and click “Add to VIPs”
Peace out,
Daily Zaps Team
How much did you enjoy this email? |