- ThePrompt
- Posts
- New AI wave with SAM 🌊
Hi folks!👋🏻 This is The Prompt! We're your go-to source for the latest AI trends!
Today, we feature SAM and the projects that are starting to be built around that model. Maybe you'll get some inspiration!
+ We take a look at some of the most interesting AI startups from Y Combinator's latest demo day.
Let's get it 👇
FEATURED
New AI wave with SAM 🌊
SAM (Segment Anything Model) by Meta, which we covered in our last issue is starting a new AI wave.
Folks are combining it with other models to create a bunch of amazing things.
I wanted to highlight two interesting projects that might give you inspiration for your own idea/project/demo to play with!
Grounded SAM model 🦕
SAM is great at segmentation, but it needs prompts to generate masks. Grounding Dino is a model that can generate high-quality masks with texts.
Combine the two and you have a model to detect and segment everything with text.
Add Stable Diffusion in the mix, and you can “edit” objects in the photos.
An interesting use case for that: Change hairstyles with text.
Edit Anything by SAM
This project aims to Edit and Generate Anything in an image.
It combines ControlNet, BLIP & Stable Diffusion so far.
Blip2 prompt: “A black drone flying in the blue sky“
What can you build?
It looks like the next phases of these projects are: ChatGPT integration, interactive controls & automatic editing, and labeling of photos.
Hopefully, this gave you some inspiration!🙌🏼
WHAT ELSE IS GOING ON
🦙 NY Times made the AI-generated images Turing test. See if you can find the real photo vs the AI-generated ones.
👀 The theme for the latest YC demo day was — you guessed it: AI! And most of the startups said that they launched a few weeks ago and that have already a good ARR. The AI bubble is here.
🏋🏻♀️ Wanna fake a sick day? Well, AI can catch you in the act. Researchers built an AI that can tell if you’re sick just from your voice.
RESOURCES
The best resources we came across lately that will help you become better at writing prompts & building AI apps.
📚 Trend in AI: Models all the way down (Stacking) [ Twitter Thread ]
👋🏻 SceneDreamer: 2D to 3D Hugging face demo [Free demo]
🎥 The LLaMa effect for open-sourced models [Interesting article]
TOOLBOX - YC EDITION
The most interesting AI tools from the latest YC demo day!
Type – AI-first document editor.
Iliad – Generate game art assets.
Layup – Build workflows across apps with one line command, like onboarding a hire.
Nucleus – AI-powered onboarding orchestration that understands “the true nature of a business.”
Hadrius – SEC-compliance robo-advisor.
Speedybrand – Generated marketing content for SMBs.
Quazel – Language learning with an AI tutor.
Booth.ai – Generative AI “photographer” for e-commerce.
Squack – Natural language accountant tools.
Berri.ai – Creating ChatGPT apps as a service.
Semantic – Financial news insights “enriched” by AI.
Credal.ai – ChatGPT-like interface for employees that references company docs but protects business secrets
Defog – Add AI data assistant to your app.
Linkgrep – AI-browser copilot for teams
Sail – Automated sales emails.
Aiflow – Automate market research based on reviews and feedback
Tennr – Turn knowledge base into a custom LLM.
Truewind – AI powered bookkeeping and finance processes.
Flair labs – Collect insights from customer service calls and email
Justpaid – Automate bill pay, catch over-payments to vendors.
Kyber – Automate insurance industry tasks like answering questions and underwriting.
Meru – Platform for training your own LLMs.
Sameday – AI that calls workers like plumbers and roofers to make appointments
Zenfetch – Analyze customer calls live and surface talking points.
Syncly – AI to analyze customer emails.
Pair AI – Video courses generated using AI.
Latent – Automating electronic health records.
Avoca – AI receptionist to answer missed calls at SMBs.
PROMPT OF THE DAY
TOOL
Midjourney V5PROMPT
A Pixar movie about a demon boy who lives in Hell, his father is Satan, and he finds himself on Earth befriending a human girl.RESULT
LATEST PAPERS
InstantBooth: Personalized Text-to-Image Generation without Test-Time Finetuning
Generative Agents: Interactive Simulacra of Human Behavior
SparseFormer: Sparse Visual Recognition via Limited Latent Tokens