- The AI Plug
- Posts
- #28 A human must launch nuclear weapons
#28 A human must launch nuclear weapons
AND: Prison for AI-generated child pornographer
The AI Plug
April 29, 2023
A bit of a different schedule this week, guys; apologies for that. I’m travelling between the UK and Europe, returning home next week - so this’ll be our last issue until next Wednesday.
Nonetheless, this issue talks about AI in nuclear war and AI child pornography - two pretty gnarly topics we should all care about.
So without further ado, let’s dive in.
In today’s newsletter:
Top News: AI shouldn’t launch nuclear weapons, NVIDIA releases guardrail software, and prison for AI child porn creator
The Rundown: 16% staff layoffs, police officers use less excessive force, and new Microsoft AI-powered tool
Tips, Tricks & Tools: Free image enlarger, digital marketing solutions, and website ranking
Hot Chat 3000: Chat 1-on-1 with other users, according to your AI-assigned hotness
Anyone and everyone: All are welcome in the AI industry
Top News 🔝
Three biggest stories if you’re in a rush
“Any decision to launch a nuclear weapon should not be made by artificial intelligence.”
We know we're living precariously when this sentence is written in an American Department of Defence policy. As announced earlier this week, ADD representatives introduced the Block Nuclear Launch by Autonomous AI, which states that no autonomous system without meaningful human oversight “can launch a nuclear weapon or select or engage targets” with the intention of launching one. To reinforce already existing Pentagon regulations mandating human action in any nuclear weapons decision, the 2023 US bill is introduced to encourage Russia and China towards similar commitments.
- - - - - - - - - - - - - -
As industries across the board adopt generative AI powered by large language models (LLMs), safety and aligning these systems with human moral code is a prevailing concern. With their recent NeMo guardrail release, NVIDIA is helping developers build impressive generative AI models which stay accurate, appropriate, secure and within the company’s field of expertise. Importantly, NeMo is an open-source model, meaning the original code is freely available for use and redistribution - thus encouraging other LLM developers to adopt similar safety precautions.
- - - - - - - - - - - - - -
Canadian Steven LaRouche, 61, has pled guilty to creating AI-generated footage of child pornography. In the country’s first-ever case involving deepfakes of children being sexually exploited, judge Benoit Gagnon stated that despite no children being harmed in the creation of the footage, children whose likeness was used to generate the content had their sexual integrity violated. The ruling comes alongside a recent California legal act, which attempts to criminalize AI-generated porn without consent.
The Rundown 🐂
Deeper dive into trending stories
US senator introduces bill to address AI-regulatory shortfalls. (link)
ChatGPT potentially hinders the AI industry. (link)
Avatars as Actors: Will AI lead to celebrity simulation rights? (link)
MediaTek, Taiwan’s top chip design company, rapidly re-focuses on AI. (link)
Police officers use 36% less excessive force thanks to AI technology. (link)
Pinecone, a vector database startup, drops $100M on $750M evaluation. (link)
Microsoft’s AI-powered designer tool is available in preview. (link)
Hot Chat 3000 uses AI to rank your hotness and determines who you can chat with. (link)
NSA cybersecurity director: “Buckle up for generative AI.” (link)
16% of Dropbox’s staff are laid off thanks to AI. (link)
Celebrities all sing each other’s songs on AI-generated tik tok reel. (link)
“AI” said over 50 times on Google, Meta and Microsoft’s earnings calls. (link)
Tips, Tricks & Tools ⚒️
Discounts, trials and productivity-boosting hacks
Murf AI - High-quality, natural-sounding AI voices from text input for your projects. (link)
Markopolo AI - AI-powered digital marketing solutions - compelling. (link)
AI Image Enlarger - FREE online image enlarger to upscale and enhance small images immediately. (link)
Longshot AI - Comprehensive SEO-content generation solution - recent topics, engaging articles, rank your website! (link)
quest AI - Powerful website frontend developer - build a beautiful website from an image in minutes. (link)
Hot Chat 3000
Upload a photo, get judged, chat with your equals (according to AI)
From their manifesto:
“HotChat 3000 is built from pre-existing stock parts: It is a snapshot of what a convenient assemblage of publicly available models and datasets determine human hotness to be. HC3K visualizes a system we are or will soon be subjected to.
There are no good publicly available models “intended” for rating human attractiveness (a use case with obvious moral transgression), but this is irrelevant: the AI space dreams of general purpose function. Necessarily, general-purpose models are for attractiveness rating, in the same way that they are for conversation, code generation, search, data processing, etc. In effect, every obviously “bad” use of AI is included in the aspirational scope of every general-purpose AI project.”
Whatever Hype Train is Leaving The Station
It’s funny because it’s too true
How’s Today’s Timing?
What’re your thoughts on a weekend issue?
With a disruption to our publishing schedule, we’ve had to push this issue out a day late.
Did you like it? Would you prefer we keep Monday, Wednesday, Friday, or do you like getting an issue on the weekend? Let us know below - we’ll make changes depending on the majority vote.
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