What's the situation with AI water usage? It turns out it's kinda complex.

We're looking at this Wired article:

https://www.wired.com/story/karen-hao-empire-of-ai-water-use-statistics/

It examines the controversy and complexities surrounding the water consumption of Artificial Intelligence (AI) and data centers. It discusses how exaggerated statistics about AI water use, such as a major error found in a journalist's book concerning a Google data center in Chile, have been challenged by critics, including an individual who asserts the severity of the issue is "fake." The article explains that data centers primarily use water for cooling processors, and the actual consumption varies greatly depending on location, climate, and technology used, with some companies opting for recycled water. Furthermore, it argues that concerns about AI's water footprint, especially when compared to other thirsty industries like agriculture and golf courses, are often rooted in a broader societal rejection of AI's rapid growth and perceived value. Ultimately, the text advises that while some water estimates may be flawed, the environmental impact of data centers cannot be ignored in water-stressed regions, highlighting the need for transparency from technology companies.

#artificialintelligence #aiethics #ai #environment #science

Hosted on Acast. See acast.com/privacy for more information.

Let's discuss a proposed paradigm shift in how we view and interact with large language models (LLMs), moving away from seeing them as conscious entities with opinions and toward understanding them as simulators.

The tweet in question: https://x.com/karpathy/status/1997731268969304070

This concept, initially highlighted in an Andrej Karpathy tweet, suggests that users should prompt LLMs to simulate diverse perspectives or groups of people rather than asking for the LLM's non-existent personal thoughts, thereby producing richer outputs.

We will also talk about the wide-ranging implications of this simulator framework, and its profound effects on technological development, philosophical understandings of agency, ethical governance, and ultimately, how humans will interact with and rely upon artificial intelligence in the future. The core message across the texts is that LLMs are powerful tools for reflecting collective human knowledge through statistical simulation, not artificial minds.

#artificialintelligence #consciousness #airesearch #ai

Hosted on Acast. See acast.com/privacy for more information.

Today's episode is about the viral dance clip featuring actor Jon Hamm in the 2025 series Your Friends & Neighbors, contrasting it sharply with his former restrained character, Don Draper from Mad Men.

We analyse how the clip—showing Hamm's character, Coop, dancing with uninhibited freedom—has become an ubiquitous internet meme representing escapism and dissociation.

Specifically, the meme is interpreted as an avatar for modern burnout and stress, allowing users to humorously express the feeling of being mentally elsewhere while dealing with overwhelming real-life responsibilities.

This juxtaposition of the stressful scenario with Hamm's absurd, carefree movement captures a survival humour common among younger generations who use irony to process anxiety. Ultimately, we think that the meme’s success lies in its visualisation of the private fantasy of psychological release from the pressures of contemporary adulthood.

The music track often accompanying this scene:

Kato Feat. Jon - Turn The Lights Off [Official Video] HD

https://www.youtube.com/watch?v=FxDMMeWK1ZE

#jonhamm #memes #viral #madmen #dondraper

Hosted on Acast. See acast.com/privacy for more information.

For this episode let's challenge the restrictive view of creativity propagated by those who demand artists “pick up a pencil” when responding to the discussion of AI-generated works.

We definitely characterise this minimal stance as “gatekeeping” and an ironic defence of a self-imposed aesthetic comfort zone, arguing that true art has always been about rebellion, experimentation, and a vast universe of practice far beyond simple drawing.

To illustrate this expansion, we invoke the radical precedents set by 20th-century artists such as Marcel Duchamp and Marina Abramović.

We think these historical figures would find the modern medium debate baffling, asserting that the idea is the art or that the experience itself transcends the instrument. Ultimately, we conclude that technology serves not to kill existing artistic practice but to expand its frontier, distinguishing those curious about the future from those defending nostalgia disguised as morality.

#aiart #antiai #aidebate #artificialintelligence #ai #aiethics

Hosted on Acast. See acast.com/privacy for more information.

Today we're looking at "adversarial poetry", a method to jailbreak AI.

The paper: https://arxiv.org/html/2511.15304v1

The academic paper and news report detail the discovery of adversarial poetry, a "universal single-turn jailbreak" mechanism that exploits a fundamental flaw in Large Language Model (LLM) safety alignment. This technique involves reformulating harmful requests into poetic verse, which successfully circumvents refusal mechanisms across 25 frontier models, including offerings from Google, OpenAI, and Anthropic.

The research found that using poetic framing led to a sharp increase in Attack Success Rate (ASR), jumping from an average of 8.08% to 43.07% for systematically transformed prompts. This stylistic manipulation was effective across numerous hazard categories, such as cyber-offense enablement and harmful manipulation, indicating that safety filters are narrowly optimized for conventional, prosaic language. Furthermore, the study revealed an inverse correlation between model size and robustness in some families, with certain smaller models displaying unexpected resilience against the poetic prompt attacks.

Additional sources:

https://www.planksip.org/platos-critique-of-poets-and-artists

#adversarialpoetry #artificialintelligence #ai #jailbreak #technews #technology

___

What do you think?

PS, make sure to follow my:

Main channel: https://www.youtube.com/@swetlanaAI

Music channel: https://www.youtube.com/@Swetlana-AI-Music

Hosted on Acast. See acast.com/privacy for more information.

Let's talk about Nano Banana Pro, which Google released recently, along with Antigravity and Gemini 3.

First things first, my book "Learning to Cope With AI" is here:

https://www.amazon.com/dp/B0DLVDMFC4

Furthermore:

https://antigravity.google

and

https://blog.google/technology/ai/nano-banana-pro/

Nano Banana Pro (Gemini 3 Pro Image), Google's advanced AI model for image generation and editing, is powered by the Gemini 3 Pro language model. This new tool significantly improves on previous versions by offering accurate, multilingual text rendering within images and supporting high-resolution outputs up to 4K. The documents highlight its integration into various Google products, including the Gemini app, Google Workspace, and the Google Antigravity developer platform, where it is used by coding agents to generate UI mockups and visual assets. Professional and creative users gain studio-quality controls over elements like lighting and camera angle, while all generated images are embedded with a SynthID digital watermark for transparency and verification. Though more expensive and sometimes slower than its predecessor, the model is praised for its ability to create complex, factually grounded visuals through Google Search integration and maintain character and brand consistency.

#nanobananatutorial #nanobanana #gemini #ai #artificialintelligence #nanobananatrend

___

What do you think?

PS, make sure to follow my:

Main channel: https://www.youtube.com/@swetlanaAI

Music channel: https://www.youtube.com/@Swetlana-AI-Music

Hosted on Acast. See acast.com/privacy for more information.

In this episode of Swetlana AI podcast we're discussing the Butlerian Jihad, a significant historical event within the Dune universe. This conflict, also known as the Great Revolt, spanned from 201 B.G. to 108 B.G. and culminated in the destruction of all thinking machines and conscious robots across the known universe.

https://dune.fandom.com/wiki/Butlerian_Jihadvfghtjzkuli

We're also touching upon the causes of the rebellion, stemming from a technocratic class controlling humanity via advanced machines, and outlines the profound legacy of the Butlerian victory, which includes a total technological reversal and the establishment of the central religious precept: "Thou shalt not make a machine in the likeness of a human mind."

The resulting technological gap led to the rise of specialized human orders like the Mentats, Bene Gesserit, and the Spacing Guild, shaping the subsequent feudal-arranged galactic order under House Corrino. Finally, the text explores the Behind the Scenes literary function of the Jihad, noting that author Frank Herbert used it as a plot device to focus on social and philosophical issues rather than technological ones, drawing parallels to works like Samuel Butler’s Erewhon.

#Dune #ButlerianJihad #DuneExplained #FrankHerbert #SciFiLore #duneuniverse

___

What do you think?

PS, make sure to follow my:

Main channel: https://www.youtube.com/@swetlanaAI

Music channel: https://www.youtube.com/@Swetlana-AI-Music

Hosted on Acast. See acast.com/privacy for more information.

This episode is a detailed profile of Steven Cheung, the White House Communications Director appointed by Donald Trump in 2025.

First we have a short biographical overview of Cheung's life, education (including his time at California State University, Sacramento, without earning a degree), and political career, highlighting his work on various Republican campaigns, his previous role in the Trump administration, and his background with the Ultimate Fighting Championship (UFC).

This article is key:

https://www.theatlantic.com/politics/archive/2025/03/steven-cheung-voice-trump/682211/

The Atlantic article and the accompanying psychological analysis focus on Cheung’s "weaponized communications" style, describing it as relentlessly aggressive, profane, and focused on public humiliation, mirroring Trump’s own approach. This style, influenced by the UFC ethos of dominance and spectacle, is characterised as a strategic system designed to colonise perception and erode moral boundaries, making Cheung the effective "voice of Trump" in his second administration.

Additional info: https://en.wikipedia.org/wiki/Steven_Cheung

#stevencheung #uspolitics #trumpadministration #politics #whitehouse

___

What do you think?

PS, make sure to follow my:

Main channel: https://www.youtube.com/@swetlanaAI

Music channel: https://www.youtube.com/@Swetlana-AI-Music

Hosted on Acast. See acast.com/privacy for more information.