The AI discussion thread

[jpiniero]

https://www.reddit.com/r/ChatGPT/comments/1nwfxnk/prompt_golden_retriever_is_pulled_over_by_police

 

[Kaido]

 

[Kaido]

Showcase of the week:

 

[igor_kavinski]

GitHub - hacksider/Deep-Live-Cam: real time face swap and one-click video deepfake with only a single image

real time face swap and one-click video deepfake with only a single image - hacksider/Deep-Live-Cam

github.com

Teens and old people are so EFFED!

 

[Kaido]

@RossMAN Hot Deals AI!!

https://twitter.com/x/status/1974685980302123213

 

[RossMAN]

@RossMAN Hot Deals AI!!

https://twitter.com/x/status/1974685980302123213

I knew this day was coming ...

 

[ArrogantHair]

Anyone see this post on the bird:

https://twitter.com/x/status/1974015430886117763

 

[Kaido]

Another showcase...you can truly make ANY image you want today!

 

[Kaido]

Midjourney photo-film emulation is getting quite good:

https://twitter.com/x/status/1964418486559560116

https://twitter.com/x/status/1928424652881232035

 

[Kaido]

Anyone see this post on the bird:

https://twitter.com/x/status/1974015430886117763

DosLLM, baby!

 

[Kaido]

Draw & prompt!

AI Image Generator & Photo Editor – Edit Images Free Online

Next-Level AI Image Editor and Generator on a infinite canvas. Runs online on both desktop and mobile. Seamlessly edit and expand images. Try for free now

visuali.io

Pure magic lol:

https://www.reddit.com/r/AgentsOfAI/comments/1nz25z8/this_is_what_art_looks_like_when_human_and_ai

 

[Kaido]

2-hour "intro to AI" video:

https://twitter.com/x/status/1974218904722681990

Free Microsoft AI boot camp: (12 weeks, 24 lessons)

GitHub - microsoft/AI-For-Beginners: 12 Weeks, 24 Lessons, AI for All!

12 Weeks, 24 Lessons, AI for All! Contribute to microsoft/AI-For-Beginners development by creating an account on GitHub.

github.com

https://twitter.com/x/status/1974062560769925441

 

[Kaido]

11-minute deep-dive on Sora 2:

 

[Kaido]

ElevenLabs UI: Open-source components for AI audio & voice agents.

https://twitter.com/x/status/1975622003550802141

 

[RnR_au]

 

[Kaido]

Youtube comment:

The future isn’t Skynet, it’s mid-tier SEO optimization.

 

[balloonshark]

A swing and a miss from Copilot on this question lol.

 

[Kaido]

Static image input + Video input

+ Relighting from picture
+ Special effects from picture

https://twitter.com/x/status/1976259706214592747

 

[Kaido]

From the thread start 2.5 years ago to today:

https://twitter.com/x/status/1976042197154963702

 

[Red Squirrel]

A swing and a miss from Copilot on this question lol.

It's accounting for data compression!

 

[Brainonska511]

https://bsky.app/profile/junoryleejournalism.com%2Fpost%2F3m2qdqsffuk2w

The creator of the wire is not a fan of the LLMs.

 

[RnR_au]

uhoh...

It reveals a surprising finding: in our experimental setup with simple backdoors designed to trigger low-stakes behaviors, poisoning attacks require a near-constant number of documents regardless of model and training data size. This finding challenges the existing assumption that larger models require proportionally more poisoned data. Specifically, we demonstrate that by injecting just 250 malicious documents into pretraining data, adversaries can successfully backdoor LLMs ranging from 600M to 13B parameters.

https://www.anthropic.com/research/small-samples-poison

 

[Kaido]

NO NOT AI PIZZA RAT

https://twitter.com/x/status/1976256674051785110

 

[Kaido]

Whoooaaaaa this has just democratized animation!!

With LTX Studio, you can generate stylized animated content using prompts, presets, or reference images. Then build a full shot sequence.

LTX Studio is your full creative pipeline.From style to motion, layout to sequencing, you can build entire animated projects in one place.

1. Choose your animation style:

LTX Studio includes built-in animation presets like:

• Anime
• Hand-drawn
• 3D
• 2D Novel

These allow you to quickly lock in a consistent style.

2. OR define your style with a prompt or a reference:

Write a clear prompt that sets the scene. Include subject, environment, lighting, and most importantly, the art style you want the animation to match.

Have a specific visual style in mind? Upload a reference and generate new shots that match the tone, lighting, and composition.

3. Use Multi-reference to create consistency across shots:

Bring together multiple image references to generate a series of shots that share the same look and feel.Perfect for building out your sequence with visual continuity.

4. Add your shots to the Storyboard and animate:

Once your key frames are ready, drop them into the Storyboard.

From there, you can add motion prompts, control the camera, and animate your shots into a full sequence.

https://twitter.com/x/status/1976319282062688321

 

[manly]

TIFU

Climate Goals Go Up in Smoke as US Datacenters Turn To Coal - Slashdot

US datacenters are experiencing a significant shift toward coal-powered energy due to elevated natural gas prices and rapidly growing electricity demand. From a report: According to a research note from financial services firm Jefferies, datacenter operators are racing to connect new capacity to...

news.slashdot.org

But Dot Com Bomb 2.0 is incoming...

UK's Central Bank Warns of Growing Risk That AI Bubble Could Burst - Slashdot

The Bank of England has warned there is a growing risk of a "sudden correction" in global markets as it raised concerns about soaring valuations of leading AI tech companies. From a report: Policymakers said there were also threats of a "sharp repricing of US dollar assets" if the Federal...

news.slashdot.org

Without data centers, GDP growth was 0.1% in the first half of 2025, Harvard economist says | Fortune

Is a U.S. without data centers a country without GDP growth?

fortune.com