#69 AI-driven Valkyrie jets

Plus: Is this our 'Terminator' moment?

AI has now started pulling the trigger, dictating target choices and even neutralizing threats. Today’s newsletter doesn't just give you news; it offers a window into tomorrow. From budgetary moves for AI-driven combat to the catch-22 of "open" AI, and from a newfound power for social media users to sidestep attention-seeking algorithms, we've captured it all for you. Let’s dive in!
 

In today’s newsletter:

  • Hottest stories: AI Valkyries, closed source models, and opt out of attention-grabbing algorithms

  • TL;DR Rundown: AI-writing AI’s, best 5 AI movies, and this is how we might know it’s conscious

  • Tool of the day: Great photos from just a single headshot 

  • Issue 69: Special meme and art for milestone issue!

HOTTEST STORIES 
Today’s biggest stories if you’re in a rush 

The US Air Force is eyeing a $5.8 billion budget slice for AI-backed combat aircraft, spotlighting the XQ-58A Valkyrie. This unmanned aerial vehicle aims to be a trusty sidekick to manned jets, beefing up firepower and support. With AI at the helm, the Valkyrie might soon autonomously strategize its moves. A simulation is on the cards this year, where it'll plot its own course to pursue and neutralize a Gulf target.

This budget bid, awaiting Congress' nod, maps out a five-year game plan centered on AI-driven combat craft like Valkyrie. This move isn't out of the blue; the Valkyrie has been in the skies, connecting dots for F-22s, F-35s, and the Air Force's Skyborg initiative.

While many in the military are giving thumbs up to this next-gen air dominance push, there's a ripple of unease. Some worry this could edge us closer to a darker, "Terminator"-esque future. As Mary Wareham from Human Rights Watch points out, we might be blurring ethical boundaries when we let algorithms, not humans, call life-or-death shots.

- - - - - - - - - - - - - -

Researchers spanning Carnegie Mellon University, the AI Now Institute, and the Signal Foundation have shed light on a surprising twist in the AI world. Even models dubbed "open" might have some strings attached.

Case in point? Llama 2 by Meta. Sure, you can download, tweak, and roll it out, but it's not your usual open-source fare. You want to use it to train other models? Think twice. Planning to integrate it into an app that's got over 700 million daily users? Better get that special license.

With these clauses, Llama 2 can easily become Meta's secret weapon, letting them cash in on enhancements made by external devs. Now, compare this with GPT Neo by EleutherAI, which sports a traditional open-source tag. Sounds great, but here's the catch: such models struggle to level up. Why? Advanced training data's often under wraps, and the tech blueprints? Dominated by tech giants.

- - - - - - - - - - - - - -

In the European Union, a digital wind of change is empowering internet users, granting them unprecedented command over their social media experience.

The Digital Services Act (DSA) is the catalyst, giving users of giants like Facebook, Instagram, TikTok, and Snapchat the reins to confidently opt out of AI-curated "personalized" feeds. No longer are users bound to the whims of algorithms deciding what's "relevant" based on tracking; instead, they can relish a straightforward, chronological feed filled with updates from friends.

This seismic shift, born from the EU's DSA, underscores a pivotal moment of user agency, where people can dodge the often overwhelming pull of attention-driven algorithms. Interestingly, while this regulation is EU-centric, its ripples are felt globally. Many tech behemoths, preferring uniformity, are rolling out these user-centric changes worldwide.

A prime example? Facebook’s proactive global embrace of a chronological news feed, entirely devoid of the "Suggested For You" posts. It's a move that, while influenced by the DSA's mandate, epitomizes the era of users regaining control, sidestepping the maze of tracking-based content suggestions.

TL;DR RUNDOWN
Listicle of what else is happening today 

Ali ali: Alibaba unveils two innovative AI models, Qwen-VL and Qwen-VL-Chat, elevating image understanding and enhancing conversational depth.

Modular AI: A trailblazing AI startup, secures a robust $100 million in funding, with heavyweights like General Catalyst and GV (Google Ventures) backing its vision.

Movie night, anyone? Cointelegraph lists five must-watch AI-themed movies that explore the intricate bond between humans and machines.

Llama writers: Meta unveils Code Llama, a cutting-edge tool powered by its Llama 2 model, destined to both craft new code and refine existing human efforts.

BioGPT: Pioneering minds behind Meta's bio-focused AI language model now steer EvolutionaryScale, a fresh startup that's already garnered a striking $40 million in funds.

Bitcoin AI: The melding of AI and blockchain is poised to reshape the music realm, amplifying artistic prowess and championing transparent, equitable earnings for creators.

How will we know? Some possible answers: A consortium of 19 experts spanning neuroscience, philosophy, and computer science introduces a checklist, potentially pinpointing systems with genuine consciousness potential.

TOOL OF THE DAY 
AI tools we’ve used, loved, and recommend above all others 

AIportraits.org is a really fun AI tool that allows you to transform your selfies into stunning works of art. With the latest AI technologies, you can choose from dozens of portrait art styles based on genuine masterpieces of all time. Ever wanted to be one of those people portrayed in the most famous paintings? Now you can, with AIportraits.org.

It’s easy and fun to use, so why not give it a try and see yourself in a whole new light? 😊

ISSUE 69  
Memes and art to celebrate a milestone 

Actually heartbreaking 😂

AI art - Spambots

Could You Help Us Out?
Share with one AI-curious friend and receive our in-depth prompt guide. Use this link

How’d we go today? Telling us makes a huge difference, it really does.