- The AI Plug
- Posts
- #30 The Whitehouse Summons Them
What a week - looking forward to a weekend of relaxation!
But before we can do that, Kamala Harris has words with leading AI-firms CEOs, Bing AI is available to everyone, and we’ve got a resource for you that is most SHOCKING.
Na just kidding with that clickbait..the resource is pretty cool though.
So without further ado, let’s dive in.
In today’s newsletter:
Top News: Harris summons leading AI CEOs, Bing AI open for everyone, and Sam Altman is possibly looking for $100B in funding
The Rundown: Meta warns of ChatGPT scams, AI supply chain startup secures #30M in funding, and AGI posible in a few years.
Tips, Tricks & Tools: Hands-off website builder, image background remover and draft seriously human emails
The AI Timeline: A really cool resource showing the past several years of AI, and how we’ve got to where we are
Trekkers knew best: Did Star Trek predict the end of humanity?
Top News 🔝
Three biggest stories if you’re in a rush
In one of the heaviest “student to the principal’s office” moments ever, Vice President Kamala Harris summoned CEOs of companies leading the AI charge, such as Microsoft, OpenAI, Anthropic and Alphabet (Google’s parent), to The Whitehouse earlier this week. During the meeting, Harris reiterated the “ethical, moral and legal responsibility” these private firms have, and recognized that while AI can address some of the biggest challenges facing America and humanity, it can also “be used to violate rights, create distrust and weaken faith in democracy.” The move by the Biden administration shows that despite their funding of AI research, they are also committed to getting more safeguards for AI.
- - - - - - - - - - - - - -
Before now, users wanting to try Microsoft’s Bing Chat AI had to sign onto a waiting list, and hope they were selected (I know, I signed up, but never heard back :( ). As of yesterday, Microsoft’s upgraded, plug-in supporting Bing AI chatbot is now entirely available to anyone who signs into the new Bing or Edge with their Microsoft account. Bing has been running on limited access since its launch in February, and it now seems like Microsoft is happy enough with where it’s at. Let’s see what types of new hallucinations we get on a mass scale.
- - - - - - - - - - - - - -
Running large language models (LLMs) isn’t cheap nor for the faint of heart. During the development of ChatGPT, OpenAI reported losing a total of $540M in 2022. While yearly user-generated revenue for GPT is projected to reach hundreds of millions, operating costs are also growing rapidly for the company. This leads Sam Altman to allegedly say he will be looking to raise over $100B in the next few years to support the development of AI “so smart that it can improve itself.”
The Rundown 🐂
Deeper dive into trending stories
“Some form of AGI” is possible within a few years, says Google DeepMind CEO….yippee. (link)
Autonomous scientific robots could run one million microbial studies a year, potentially driving a significant leap forward in medical and environmental research. (link)
AMB is expanding into AI chips, and Microsoft is helping finance it. (link)
Meta warns users that it’s seen a sharp uptick in malware disguised as ChatGPT and other generative AI software, saying ChatGPT scams are the new crypto scams. (link)
Pando, an AI-powered supply chain startup, secures a $30M investment. (link)
Slack updates look to put AI at the center of their user experience. (link)
Salesforce and Accenture partner up to help clients harness generative AI in a new acceleration hub. (link)
Google expands their testing base for generative AI in Gmail and Docs by 10-fold. (link)
The AI race is disrupting education firms, leading to potentially harmful consequences for the sector. (link)
LinkedIn just turned 20 years old and celebrates its birthday with a big AI push. (link)
Explore the discord channel where thousands of rogue AI producers are making (great) AI music. (link)
General mindset, alignment research and governance, the three tenants of AGI safety. (link)
WellSpan Health, an AI used in Susquehanna Valley hospitals, helps medical professionals find abnormalities in medical scans. (link)
TIPS, TRICKS & TOOLS ⚒️
There’s an AI for That - Database of over 24,000 AI tools for 578 different tasks. Used by over 500,000 people. (link)
Magical AI - ChatGPT-powered Chrome extension helps draft emails or messages seamlessly anywhere online. So easy to use - strongly recommended. (link)
Durable AI Website Builder - Build a unique website in 30 seconds using AI. (link)
Remove.bg - Remove any image background for free using AI (a premium Canva feature - nice, it’s free here). (link)
Delve AI - SaaS solutions enabling exceptional customer service and rapid business growth. (link)
AI/ML/LLM Transformer Models Timeline & List ⚒️
A most enlightening look into the recent history of AI development
Our modern-day AI is very, very sophisticated. AI has indeed come a long way in a very short time, and what an expensive and long journey it’s been.
As Viktor Garske’s resource shows (you can see it properly here), the first version of GPT went live in mid-2018. While impressive, this model was far inferior to GPT2, which launched just nine months later in early 2019.
The models grew in complexity and capability, until the recent launch of GPT3, 3.5 and 4, which have arguably been the most significant fuel for the current AI boom.
But it’s been so much more than that - take a look at the full timeline above to get an idea of the scale of development which has occurred in AI over the past several years.
Help Us With Our Sources
We really need your input
Choosing reliable, authoritative sources is vital. We need to know that the news we share with you comes from verifiable, legitimate sources.
Sometimes, this means we need to link to gated content - often from large news sources like Reuters, Bloomberg, and the NY Times.
Unless you have subscriptions to these sites (most people don’t), you won’t be able to read the original stories if we link to them.
This leads to the question:
As Always, Star Trek Knew It
I wonder if Elizier is a trekker..
So this is exactly how Elizier Yudkowsky sees AI playing out for humanity - that it develops to such a level that it recognizes how inefficient we humans are, and destroys us to optimize its environment.
Imagine that…if that’s the reality we live in, one where Star Trek accurately predicted the end of all life.
Spread The Word
Share and earn epic rewards
Randy’s gonna help us - will you?
Tell one person about our new newsletter design by sending them the link below ⏬.
If you do, we'll send you a totally free, extensive course on prompt engineering, where you will learn to:
Get to most out of GPT and Midjourney
Automate workflows
Free up your time
Do more cool sh*t instead of working