- WeeklyDispatch.AI
- Posts
- The week in AI: OpenAI launches the GPT Store & ChatGPT for teams
The week in AI: OpenAI launches the GPT Store & ChatGPT for teams
Plus: AI steals the 2024 Consumer Electronics Show

Welcome to The Dispatch! We are the newsletter that keeps you informed about AI. Each Thursday, we aggregate the major developments in artificial intelligence; we pass along the news, useful resources, tools or services, and exciting projects in open source. Even if you aren’t an engineer, we’ll keep you in touch with what’s going on in AI.

OpenAI has announced two major updates to the ChatGPT landscape this week. The GPT Store rolled out for most ChatGPT Plus and Enterprise users on Wednesday. GPTs are custom versions of ChatGPT with user-defined operating instructions, extra knowledge and any combination of skills. With the GPT Store, users will be able to browse the top/featured models and GPT creators will be paid. |
And it’s pretty easy to create a GPT. It appears likely that domain expertise will be the dominant use case for GPTs - but there are GPTs for everything from guided journaling to personalized shopping assistance. GPTs are still quite new, so only time will tell how useful and popular these customized models become and what kind of compensation their creators might receive. If you come up with a great idea for a GPT, you can apply for your GPT to be featured - OpenAI claims they review every submission.
The tech giant also announced ChatGPT for teams. For $25/month per user (billed annually; minimum 2 members to create a team), team members will receive higher message caps on GPT-4 & DALL-E usage and a dedicated workspace with admin console, team management, and shared GPT’s. OpenAI does not train on any teams data.
The World Economic Forum has named AI-powered misinformation as the greatest short-term threat to the global economy in its annual Global Risk Report. The report, based on a survey of nearly 1,500 experts and leaders, emphasized that AI could contribute to further polarizing societies and undermining democratic processes. Generative AI has made the creation of sophisticated deepfake content more accessible, and the technology can be used maliciously to fuel cyberattacks. Extreme weather was noted as the second-most pressing short-term risk, with longer-term fears focused on irreversible climate change and environmental degradation.
Are standalone AI devices poised to replace smartphones? Not quite yet - but eventually, claims AI startup Rabbit. The company’s new R1 gadget is about half the size of an iPhone and trained as a “Large Action Model” to navigate the web and use apps. In spirit, the R1 fulfills an Alexa-like role: ask for what you want, and the device delivers from a single interface. |
The R1 was among several AI highlights from Consumer Electronics Show 2024 in Las Vegas. Here are some of the others so far (event runs through Jan. 12th):
Amazon’s Alexa gets new generative AI-powered experiences
Samsung’s 2024 TVs are bigger and packed with AI and accessibility features
This smart mirror uses AI to boost your confidence and mood
Volkswagen integrates ChatGPT into its vehicles
Nvidia makes the case for the AI PC at CES 2024
Asus ROG Phone 8 Pro gaming phone with AI-focused Qualcomm Snapdragon 8 processor
A $3,500 toaster for steak: the wild AI Perfecta Grill promises sizzling meat in 2 minutes
This smart toothbrush talks through your bones to improve your brushing
Invoxia launches AI wearable that monitors your pet's health
Comedian George Carlin is the latest celebrity to be brought ‘back to life’ using AI. In the one hour special “I’m Glad I’m Dead”, Dudesy (a comedy AI platform from Mad TV alum Will Sasso and podcaster Chad Kultgen) does its best job to bring Carlin into the present for an incendiary monologue on controversial issues from Trump to transgender rights to artificial intelligence itself.
Dudesy was not granted permission for the project from the Carlin estate. In a statement, Kelly Carlin said that her father “spent a lifetime perfecting his craft from his very human life, brain and imagination” and that “no machine will ever replace his genius.”
More in AI this week:
OpenAI fires back at New York Times copyright lawsuit
Pennsylvania becomes first state in the nation to offer ChatGPT Enterprise to select state workers
AI study suggests famous Raphael painting was not entirely his own work
Google faces $1.67b damages demand at AI-related patent trial
Valve opens the door to more Steam games developed with AI
New material found by AI could reduce lithium use in batteries
China and cybercriminals are targeting American AI companies, FBI Director Wray says
Avi Schiffmann’s Tab AI necklace has raised $1.9 million to replace God
Amazon turns to AI to help customers find clothes that fit when shopping online

Trending AI tools/services:
PhotoRoom: the AI photo editing app
Fixkey: one keypress to fix typos; full context AI-correction
Genie from Luma AI: create any 3D object in under 10 seconds in all standard file formats
Gemini vs GPT: compare side-by-side prompts with Google’s Gemini Pro vs OpenAI’s ChatGPT
PocketAI: GPT-powered content creation for WhatsApp
Stripe Your GPT’s: monetize your GPT idea with Stripe using a simple one click no code template
Fliz: turn any URL into a stunning video with AI
Guides/informative/lists:
Creating AI-powered comic books with consistent characters using ChatGPT
The best robots and AI innovations we've seen at CES 2024 so far
How to access GPT-4 Turbo with Copilot for free
Notepad set to become the latest product to receive the Microsoft AI treatment
6 ways AI is transforming global health
(YouTube) Training and deploying your own open source language models
(YouTube) OpenAI GPT Store ideas + how to connect an API to your GPTs
Social media/videos/podcasts:
How AI can help turn your idea into the next great app - Replit CEO Amjad Masad at TED AI [YouTube]
The ChatGPT store is about to launch… let’s get rich [YouTube]
The Accidental AI Canvas, with Steve Ruiz of tldraw [Podcast]
Google is preparing a paid version of Bard that uses Gemini Ultra - here are some leaked features [X]
Figure-01 (AI startup Figure’s first humanoid robot) has learned to make coffee after watching humans make coffee [X]
(Discussion) Should you say please? Researchers identify 26 golden rules for prompting [Reddit]
(Discussion) Half of all skills will be outdated in two years, study suggests [Reddit]
Open source/technical/research papers:
Open Interpreter 0.2.0: platform that lets LLMs run code on your computer to complete tasks; updated with new Computer API
GPT Pilot: a true AI developer that writes code, debugs it, talks to you when it needs help, etc
AudioBox from Meta AI Research: where anyone can make a sound with an idea
An “AI Breakthrough” on systematic generalization in language?
FireAttention: serving open source models 4x faster than vLLM by quantizing with ~no tradeoffs
(Research) Mixtral of Experts: Mistral AI’s “Mixture of Experts” method for large language model that surpasses GPT-3.5 Turbo, Claude-2.1, Gemini Pro, and Llama 2 70B on human benchmarks
(Research) The Impact of Reasoning Step Length on Large Language Models: getting the most out of your Chain-of-Thought style prompts for LLM’s
Eighteen of the 20 most successful AI companies in the world are American. You can bet your bottom dollar that foreign adversaries, especially the Chinese, are actively targeting that innovation, that intellectual property.