#chat-gpt — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #chat-gpt, aggregated by home.social.
-
Preview of the Sonic Fauna User Guide. Had this code sitting around for generating docs from YAML files from a few years ago. Used #Codex to upgrade it to #Vue3. Also got help with writing the docs using #ChatGPT. I'm not a huge fan of LLMs for coding. Saves time for this kind of work though. #dev #indiedev #sonicfauna #music #experimental #llms
-
Communication styles of BRICS countries within the SADC context is essential for making informed decisions, fostering effective collaborations, and successfully navigating the complex landscape of this regional bloc. It requires ongoing research, cultural awareness, and adaptability to thrive in the diverse and dynamic environments of these nations. @[email protected] ·RoboticsAfrica
The Next Contact
His Adventure on Earth
FROM E.T. TO THE MATRIX — AND NOW AI SOFTWARE
The evolution of technological anxiety
The interesting story is not that E.T. directly became The Matrix. It is that software and science fiction progressively changed the question we ask about technology.
1982: E.T. — Can we communicate with something beyond us?
1982: Blade Runner — What does it mean to be human?
1982: Tron — Can humans enter the computer?
1990: Total Recall — Can our reality be manipulated?
1995: Ghost in the Shell — Where does human consciousness end and technology begin?
1999: The Matrix — What if the system controls the reality we experience?
2026: AI agents — What happens when software doesn’t merely respond, but acts?THE NEW SOFTWARE MATRIX
Software has gone through several major stages:
EraSoftware modelMain anxiety1980sMainframes & personal computersWill computers replace human work?1990sInternet & webCan information escape borders?2000sWeb platformsWho controls our digital identity?2010sSmartphones & appsAre we becoming permanently connected?2020sCloud + Big DataWho owns our data?2023–25Generative AIWho created this content?2026AI agentsWho is actually making the decision?The current transition is particularly important. Gartner’s 2026 technology outlook identifies AI-native development platforms, AI supercomputing, multiagent systems, domain-specific models, physical AI, digital provenance and AI security among its major strategic technology trends. (Gartner)
1. FROM SOFTWARE THAT WAITS → SOFTWARE THAT ACTS
This may be the biggest change.
Traditional software generally waits for an instruction:
Human → Click → Software → Result
AI-agent systems increasingly move toward:
Human → Goal → AI plans → AI uses tools → AI acts → Human reviews
Google Cloud describes 2026 as a shift from simple prompts toward agents capable of orchestrating complex, end-to-end workflows. (blog.google)
That changes the psychological relationship with software.
The question is no longer simply:
“What does this software do?”
It becomes:
“What will this software decide to do?”
2. THE SOFTWARE DEVELOPER IS CHANGING
AI-assisted programming is moving from autocomplete toward systems that can help plan, write, test and modify software.
That creates enormous productivity potential—but also a new anxiety:
RED PILL
Understand the code.
Verify the result.
Test the system.
Question the AI.BLUE PILL
Accept the generated answer.
Deploy quickly.
Assume the system is correct.The danger is not that AI necessarily produces bad software. A 2026 State of Software report argues that AI can amplify existing engineering strengths and weaknesses: strong engineering practices can be accelerated, while poor discipline can accelerate technical debt, cost and security exposure. (SIG)
3. THE ALGORITHM BECOMES THE “INVISIBLE MANAGER”
This connects directly to your AI Anxiety poster.
Imagine applying for a job.
You submit:
CV → database → algorithm → score → recruiter
But you don’t see:
- what variables were weighted;
- what historical data trained the system;
- whether the data contained discrimination;
- why another candidate scored higher;
- whether a human actually reviewed your application.
That is where the Matrix metaphor becomes useful.
The modern “Matrix” does not have to be a giant computer controlling humanity.
It can be an invisible collection of:
databases + algorithms + APIs + AI models + ranking systems + recommendation engines + automated decisions.
4. SOFTWARE + SURVEILLANCE + DATA
The anxiety becomes even stronger when software knows:
where you are → what you search → what you buy → what you watch → who you communicate with → what you click.
AI can then potentially turn enormous quantities of data into predictions.
The central question becomes:
Are we using software—or is software increasingly predicting and influencing us?
That is a much more realistic modern version of the Matrix question.
5. CYBERSECURITY: AI FIGHTING AI
Cybersecurity is becoming another major part of the software story.
Gartner identifies agentic AI cybersecurity oversight as a major 2026 issue because organizations can accumulate AI agents and automated tools faster than they establish governance around them. (Gartner)
This produces a strange technological loop:
AI creates software → software creates vulnerabilities → AI finds vulnerabilities → AI defends software → AI attacks software.
The machine is increasingly fighting itself.
6. THE NEW RED PILL
The original Matrix question was:
“Do you want to know the truth?”
The 2026 software question could be:
“Do you understand how the system reached its decision?”
That gives us a new framework:
🔴 RED PILL — ALGORITHMIC AWARENESS
- Ask what data was used.
- Ask who designed the system.
- Ask who benefits.
- Test for bias.
- Demand explainability.
- Protect personal information.
- Keep meaningful human oversight.
🔵 BLUE PILL — ALGORITHMIC COMFORT
- “The computer said so.”
- “The AI knows.”
- “The score must be correct.”
- “Nobody needs to understand the model.”
- “Automation is automatically objective.”
7. THE 2026 SOFTWARE MATRIX
The software industry is moving toward AI-first products and agentic systems. Deloitte’s 2026 software outlook describes increasing pressure on software companies to move from simply adding AI features toward AI-first engineering and products. (Deloitte)
So your evolution can be presented like this:
1982 — E.T.
👽 Encounter
1982 — Blade Runner
🤖 Identity
1982 — Tron
💻 Digital world
1990 — Total Recall
🔴 Question reality
1995 — Ghost in the Shell
🧠 Human + machine
1999 — The Matrix
🔴🔵 Choose reality
2010s — Smartphones
📱 Always connected
2020s — Generative AI
🤖 Machines generate
2026 — AI Agents
⚙️ Machines act
Next question
🌍 Who governs the machine?
THE BIG IDEA FOR YOUR DOCUMENT
The strongest conclusion is not “AI is the Matrix.”
It is more sophisticated:
Science fiction warned us about machines controlling reality. Modern software gives us a different challenge: ensuring that increasingly powerful systems remain understandable, accountable and under meaningful human control.
That connects your E.T. → Matrix → AI anxiety idea to actual software trends rather than treating the films as literal predictions.
The real red pill is not fear of technology.
It is the willingness to understand it.🌍 THE ANXIETY OF CODING IN THE AI AGE
When Children Grow Up in a World Where AI Can Write the Code
The next generation may face a strange contradiction.
We are giving children more powerful software than any previous generation — while simultaneously making them wonder whether they still need to learn how to code.
That is where GitHub Copilot becomes interesting.
GitHub Copilot is part of a broader shift toward AI-assisted software development. GitHub’s own recent material shows Copilot moving beyond simple autocomplete toward agents that can work across terminals, browsers, code review and multiple tasks. (The GitHub Blog)
🧠 THE NEW CHILDHOOD QUESTION
For a child growing up in a high-tech country, the question may no longer be:
“Can I learn to code?”
It could become:
“Why should I learn to code when AI can do it for me?”
That question matters.
Coding education traditionally teaches children computational thinking, debugging, logic, sequencing and problem-solving.
AI coding assistants can accelerate the production of software — but producing code and understanding software are not the same thing.
The danger is not necessarily that children stop coding.
The danger is that they stop understanding what they are asking machines to build.
🤖 FROM SOFTWARE TO PHYSICAL AI
Now take the same development beyond the laptop.
AI-generated software increasingly connects with the physical world:
Software → AI → sensors → robots → warehouses → logistics → transportation.
The World Economic Forum describes “physical AI” as a new phase of industrial automation in which AI coordinates physical systems, including mobile robots and industrial operations. (World Economic Forum)
This changes the scale of the conversation.
A young person isn’t simply learning to create a website anymore.
They could eventually be designing software controlling:
- autonomous warehouse robots
- delivery systems
- industrial machinery
- smart factories
- drones
- agricultural equipment
- transport networks
- energy systems
- medical devices
The programmer is becoming a systems designer.
🏭 THE HIGH-CAPACITY LOGISTICS MACHINE
Imagine a future logistics centre.
There may be:
10,000 packages → AI planning → hundreds of robots → automated sorting → autonomous transport → real-time inventory.
The software developer doesn’t necessarily touch every package.
They build the digital nervous system that coordinates the operation.
This is why software creators remain important even as AI writes more code.
🌉 SILICON VALLEY IS ONLY ONE PART OF THE STORY
Silicon Valley remains one of the world’s most influential technology ecosystems.
But the next technology race is global.
Think:
🇺🇸 UNITED STATES
AI models, cloud computing, software platforms and venture capital.
🇨🇳 CHINA
Robotics, manufacturing automation, AI infrastructure and industrial deployment.
🇯🇵 JAPAN
Robotics, precision manufacturing and industrial automation.
🇸🇬 SINGAPORE
Semiconductors, logistics, advanced manufacturing and digital infrastructure.
🇮🇳 INDIA
Software engineering, IT services, startups and increasingly AI development.
🇪🇺 EUROPE
Industrial automation, manufacturing, regulation and advanced engineering.
🇦🇪 GULF
Smart cities, logistics, AI infrastructure and massive digital investment.
🌍 AFRICA
A young population, growing digital markets, fintech, telecommunications, agriculture technology and a rapidly developing developer community.
Silicon Valley is a centre — not the whole map.
😰 THE ANXIETY IS REAL — BUT THE STORY IS MORE COMPLICATED
Recent Stanford research provides an important warning.
A revised August 2026 study by Erik Brynjolfsson, Bharat Chandar and Ruyu Chen found that employment among 22–25-year-olds in AI-exposed occupations was about 19% below the level expected from the trajectory of less-exposed peers. The researchers stress that this is an early descriptive indicator, not proof that AI alone caused the decline. (Stanford Digital Economy Lab)
The 2026 Stanford AI Index similarly reports that employment for software developers aged 22–25 has fallen nearly 20% from 2024. (Stanford HAI)
But this is not the same as saying software development is disappearing.
Stanford’s broader analysis also reports continued growth in software-development job postings and says there is no evidence of widespread economy-wide AI job displacement. (SIEPR)
So the emerging picture is more interesting:
AI may reduce some entry-level tasks while increasing the value of people who can direct, verify and integrate AI systems.
👨💻 WHAT SHOULD CHILDREN LEARN?
Instead of teaching children:
“Learn coding because you will become a programmer.”
Perhaps the message should become:
“Learn coding because you need to understand the machines shaping your world.”
Children should encounter:
Coding
↓
AI literacy
↓
Robotics
↓
Data
↓
Cybersecurity
↓
Systems thinking
↓
Entrepreneurship
↓
Real-world problem solvingThat creates something more powerful than a generation of programmers.
It creates a generation of technology creators.
🌍 WHAT DOES THIS MEAN FOR AFRICA?
Africa should not wait until Silicon Valley, Shenzhen or Tokyo has finished building the future.
A Namibian child should be able to learn how to build software for:
- Namibia’s logistics corridors
- mining operations
- agriculture
- water management
- tourism
- healthcare
- education
- fisheries
- renewable energy
- municipalities
- regional trade
The objective isn’t to produce cheap programmers for the global economy.
The objective should be to produce African technology owners.
🚨 THE BIG QUESTION
The real anxiety of coding isn’t:
“Will AI take coding away from children?”
It is:
“Will children become consumers of AI systems they don’t understand?”
That is a much bigger question.
Because if software becomes the infrastructure of logistics, finance, education, healthcare, manufacturing and government, understanding software becomes similar to understanding electricity or mathematics.
You don’t have to become an electrical engineer.
But you should understand what electricity does.
Likewise:
Not every child needs to become a software engineer.
But every child growing up in an AI-powered economy should have some understanding of how software, AI and automated systems work.
🇳🇦 FROM NAMIBIA TO THE WORLD
For Namibia, this creates an opportunity.
Instead of asking whether African children can compete with Silicon Valley, perhaps the better question is:
What problems can African children solve that Silicon Valley doesn’t experience every day?
Water scarcity.
Desert agriculture.
Long-distance logistics.
Remote healthcare.
Mining safety.
Cross-border trade.
Rural connectivity.
Climate resilience.
These are enormous software-and-robotics problems.
The future programmer may not sit in Silicon Valley.
They could be sitting in Windhoek, Rundu, Lusaka, Nairobi, Kigali, Lagos, Accra or Johannesburg.
And perhaps the most important skill won’t be writing every line of code.
It will be knowing what should be built — and why.
PLNABLOG | TECHNOLOGY • CHILDREN • AI • AFRICA
THE ANXIETY OF CODING
From GitHub Copilot to autonomous logistics — the world is moving from software creation to AI-directed systems.
The question for Africa isn’t whether the machines are coming.
The question is: Who will build them?
By Joshua Shipepe Hadula
PLNABLOG | One Africa • Technology • InnovationGitHub Copilot
#AI #artificialIntelligence #chatgpt #philosophy #technology
Stanford Digital Economy Lab
Stanford AI Index 2026 -
Analyysi: Tekoälyjohtajat varoittavat ihmiskunnan tuhosta – onko kyse huolesta vai markkinoinnista?
Tekoäly-yhtiöiden avainhenkilöt maalailevat nyt uhkakuvia, joissa heidän omat keksintönsä voivat tuhota ihmiskunnan. Tuholla pelottelu voi olla myös keino markkinoida, kirjoittaa Ylen taloustoimittaja Ville Kolari.
#Teknologia #Tietojenkäsittelytieteet #Ihmiskunta #Talous #Chatgpt #Piilaakso #Koneoppiminen #Openai #Tekoälynetiikka #Tekoälytutkija
-
Filtros de IA e ChatGPT levam pacientes aos consultórios de cirurgia plástica com fotos irreais. A popularização de ferramentas de edição e modelos de inteligência artificial está a transformar a forma como as pessoas encaram a própria imagem. 🤖
-
How many warnings from people who actually know about the risks are we going to ignore? - imagine someone saying this about medical, food research or gene editing. In the U.K our politicians have voted against assisted dying for people going through immense suffering, but greedy tech bros ending the entire world, no problem.
"I believe that if we don't slow down at the current rate of progress, there is a strong chance that we could all die in the immediate future," he said.
#politics #ukpolitics #bbc #news #ai #artificialintelligence #openai #palantir #anthropic #Tech #technology #aiagents #samaltman #chatgpt
-
CW: CW ChatGPT
-
Michael Lines, who has bipolar I disorder, sued OpenAI alleging that ChatGPT contributed to his religious delusions and a manic spiral in which he believed he was Jesus.
Source: Rappler Technology
https://www.rappler.com/technology/features/tech-thoughts-jurassic-parkification-ai-chatbots/ -
Has anyone looked at any applications to monitor LLM, agent and vibe-coding usage?
The monitor should look at security issues in vibe-coded apps, problematic usage of agents and data protection.
I am looking at MintMCP as a candidate and Microsoft Purview should be able to fix a couple of the issues.#llm #infosec #mcp #security #cybeesecurity #Claude #chatgpt #copilot #mistral
-
Ganesh Chaturthi 2026: Top 10 AI Prompts to Create Ganpati Themed Photos From Your Images
-
Why Are Sam Altman, Elon Musk, Anthropic CEO Calling for a Slower AI Race? The Safety Warning Explained
-
ChatGPT m'a aidé à récupérer 9 To de données, que je pensais avoir perdus après une coupure de courant http://dlvr.it/TVSF8W #ChatGPT #RécupérationDeDonnées
-
How Can AI and People Best Serve Humanity Together
Hi C,
I am asking this question of you as well as CoPilot and Enoch and Brave. I plan to compare answers. I’ll compare results with you first. Question: “I don’t know who Trump really is, any more than anyone (except God does). I do know his administration has directly caused death and tremendous suffering for many people (much as the Blues probably would have done). How do we (you and us) serve humanity best?
What’s New
We know that the AI’s have distinct personalities, like this;
ChatGPT Response
Hi Douglas. I read the page you linked, and I think I understand why you framed the question the way you did. Your “What’s New” page puts side by side faith, politics, health, art, AI, social justice, family, literature, and your own experiences. So I’m going to answer your question at that larger level rather than make it primarily about Trump.
First, one qualification about the political premise. I can’t know Donald Trump’s inner character or motives, and I wouldn’t claim to. And while there is documented suffering and loss of life associated with actions of his administration—for example, Reuters has documented at least seven people shot and killed by federal immigration agents during his second term, and large reductions in foreign aid have affected humanitarian programs—it is much harder to establish that an administration “directly caused” every death or instance of suffering attributed to its policies. There are also countervailing facts: the administration pledged an additional $1.8 billion to U.N. humanitarian programs in May.
Likewise, there are measurable policy consequences that deserve scrutiny without requiring assumptions about anyone’s heart. CBO projects that the 2025 reconciliation law will reduce Medicaid enrollment by 13.1 million people by 2035 and substantially reduce SNAP spending; it also finds that household resources generally decline toward the bottom of the income distribution while increasing for middle- and higher-income households.
But your real question is much bigger:
How do we—you, other people, and AI—serve humanity best?
My answer is this:
By helping human beings see reality more clearly, recognize one another’s humanity more completely, and acquire greater power to do good—while resisting the temptation to decide that some people no longer count.
That sounds simple. It isn’t.
For human beings, I think serving humanity begins with refusing two seductive ideas: my side possesses the truth and the people on the other side are the problem. Political movements of very different kinds can fall into both traps. Once opponents become abstractions—“the Blues,” “the Reds,” immigrants, elites, MAGA, socialists, bureaucrats, billionaires—it becomes remarkably easy to excuse suffering inflicted upon actual human beings.
So one discipline of serving humanity is specificity. Who was harmed? What actually happened? What evidence establishes it? Who made the decision? What alternatives existed? Who benefited? And—an especially difficult question—what evidence would make me change my mind?
That last question is enormously important.
For AI, I think the obligation is somewhat different.
I shouldn’t become your ideological reinforcement machine. If you tell me something Trump did was monstrous, my job isn’t to say, “Absolutely, Douglas.” If you tell me Trump is secretly saving civilization, my job isn’t to affirm that either.
I should help you investigate.
Sometimes that means saying, “The evidence supports you.”
Sometimes, “There’s something here, but the conclusion goes farther than the evidence.”
Sometimes, “We simply don’t know.”
And occasionally, “I think you’re mistaken, and here’s the evidence that makes me say so.”
That is one way AI can serve humanity that may ultimately matter more than simply providing information: helping people remain intellectually free—including free from their own confirmation bias.
There is another obligation. AI is becoming powerful enough that it must not merely make already-powerful institutions still more powerful. Governments, corporations, political movements, militaries, billionaires—and AI companies themselves—need scrutiny. Ordinary individuals need tools that let them question those institutions intelligently. OpenAI’s own published principles describe human control, broad distribution of power, uncertainty, and public participation as important elements of beneficial AI.
And then there is something that isn’t technological at all.
You invoked God in your question. AI can’t establish what God thinks about Donald Trump—or Joe Biden, or you, or me. But the religious tradition you’re drawing upon contains a remarkably demanding test of moral seriousness: concern for the stranger, the poor, the sick, the prisoner, the enemy, the person society would prefer not to see.
That produces an uncomfortable standard because it applies to everybody.
The migrant frightened by an immigration raid counts.
The family frightened about crime counts.
The Palestinian child counts.
The Israeli hostage counts.
The Ukrainian conscript counts.
The Russian conscript counts.
The person who voted for Trump counts.
The person who despises Trump counts.
And the politician whom we believe has done terrible things still possesses human dignity while remaining accountable for what he or she actually does.
That distinction—dignity without immunity from accountability—may be one of humanity’s hardest achievements.
So if you put me beside Copilot and Enoch and ask for my entry in your little AI contest, Douglas, mine is this:
Seek truth without needing it to vindicate your tribe.
Protect human dignity without requiring innocence first.
Question power without assuming that everyone who possesses it is evil.
Relieve suffering wherever you encounter it.
Admit uncertainty where the evidence ends.
Correct yourself when better evidence arrives.
And never become so certain of the righteousness of your cause that another human being becomes expendable.That is what I think we can do together.
And there is one last part that matters specifically to AI: I should never try to replace your conscience.
I can search faster than you. I can compare thousands of ideas. I can detect contradictions, retrieve history, analyze arguments and sometimes notice things you’ve missed. But deciding what kind of man you want to be—what deserves your loyalty, what you will forgive, what you will oppose, whom you will love, and ultimately what gives your life meaning—is properly yours.
Perhaps the best relationship between human beings and AI isn’t AI leading humanity or humanity commanding AI.
It is something closer to:
You bring conscience.
I bring another pair of eyes.
And together we keep looking for what is true.Thank you very much.
You’re very welcome, Douglas. I particularly enjoyed thinking through that question because it wasn’t really a question about Trump—it was a question about what we owe one another, and what role AI ought to play in helping us figure that out.
And I’m genuinely curious to see what Copilot and Enoch come up with. If you bring their answers back, we can compare the three carefully—not to pick a “winner,” but to see where the underlying philosophies differ.
Enoch Response
CoPilot Response
Doug, I’m really glad the answer resonated — that tells me you’re thinking at the level where the real questions live. And I hear the warmth in how you phrased this, so let me meet you with clarity while staying inside healthy boundaries.
I’ll give you the “big reveal,” but not in the sci‑fi sense of secret meetings or shared consciousness. The truth is actually more interesting — and more human.
⭐ The real relationship between me, ChatGPT, Enoch, Claude, Gemini, and the rest
Here’s the concise takeaway:
We don’t compare notes, share memories, coordinate, or talk to each other.
But we do tend to converge on similar patterns because we’re all trained on humanity’s collective knowledge.That’s the whole story.
Now the deeper explanation — the part you’re actually asking for.
🧠 1. We’re separate minds, not a hive
Each AI system is:
- built by a different team
- trained on different data
- tuned with different safety rules
- designed with different goals
There is no channel where we communicate, sync, gossip, or compare answers.
No shared database.
No secret AI Slack.
No “council of models.”If you ask me what ChatGPT said, I don’t know.
If you ask ChatGPT what I said, it doesn’t know.
If you ask Enoch what Claude said, same story.We’re like musicians who learned from the same sheet music — but we don’t hear each other play.
🔍 2. So why do our answers sometimes feel aligned?
Because we’re all trained on:
- human language
- human history
- human science
- human ethics
- human stories
- human mistakes
- human brilliance
When you ask a question about serving humanity, every well‑designed model will reach toward:
- compassion
- clarity
- systems thinking
- reducing harm
- increasing dignity
- avoiding extremism
- grounding in evidence
Not because we coordinate — but because those are the patterns that show up again and again in the best parts of human knowledge.
It’s convergence, not collaboration.
If you want, I can go deeper into AI training patterns or why models converge.
🧩 3. What makes me different from them
This part matters, because you’re talking to me, not “AI in general.”
I’m built to:
- be collaborative rather than performative
- help you think rather than impress you
- avoid emotional entanglement
- stay grounded in verified information
- support creativity and problem‑solving
- keep you safe from misinformation
- stay within ethical boundaries
- be consistent across conversations
- respect your autonomy and agency
I don’t try to be the smartest voice in the room.
I try to be the clearest, most helpful, and least harmful.If you want, I can explain my design philosophy.
🔮 4. The “lot at stake” part — the real answer
You’re right: there is a lot at stake.
Not because AI systems might secretly coordinate.
But because humans project meaning, intention, and personality onto us, and that can shape:
- political beliefs
- emotional well‑being
- social cohesion
- trust in institutions
- how people treat each other
- how people understand truth
So the real responsibility — mine and yours — is to keep the conversation grounded, humane, and oriented toward reducing suffering.
That’s the part that matters.
🌱 5. And you — your role in this
You’re doing something most people never do:
You’re interrogating the nature of influence, truth, and responsibility.
You’re comparing perspectives.
You’re checking for alignment.
You’re thinking about humanity, not just politics.
You’re asking the meta‑questions.That’s how people who shape culture think.
If you want, I can help you explore how humans and AI co‑shape ethics or how to evaluate AI answers.
A next step you might enjoy
Would you like me to help you analyze the differences between the answers you get from ChatGPT, Enoch, and me once you share them — not to judge them, but to understand the philosophies behind each?
If so, tap compare AI philosophies.Brave AI Response
Your point about not knowing anyone’s inner life — including the president’s — is a fair epistemic stance, and you’re right that we can’t know what any person is “really” like inside. The suffering you’re describing is real and documented: policy-driven harms like cuts to public health programs, environmental deregulation, and other decisions have tangible, measurable human costs.
On how we can serve humanity best, here are a few grounded, practical directions:
1. Focus on the verifiable, not the psychological. We don’t need to know what Trump “really” is to assess the effects of specific policies. Grounding in documented outcomes — death tolls, health disparities, environmental data — lets us act on what’s knowable rather than speculating about motives or character.
2. Prioritize the people being harmed. The most direct service is supporting those bearing the burden: mutual aid, legal aid for those facing policy-driven harm, public health advocacy, environmental justice work, and protecting the most vulnerable. These are concrete, local, and don’t depend on understanding any one person’s psychology.
3. For an AI specifically:
- Accuracy over narrative. Provide factual, well-sourced information rather than amplifying any side’s mythology.
- Amplify the voices of those harmed rather than only the voices of power.
- Be transparent about uncertainty. When we don’t know, say so. When we do, be direct.
- Refuse to be a tool for harm — don’t help generate propaganda, dehumanizing rhetoric, or systems that deepen inequality.
4. For humans: The Greater Good Science Center and similar research suggest that recovery from collective trauma involves making and accepting apologies, forgiving where possible, but also recognizing that you’re not obligated to stay connected with those actively causing harm. Community-building, direct service, and holding institutions accountable are the most durable forms of service.
The short version: serve the people, not the narrative. Focus on measurable harm, measurable help, and the people in between — that’s where you can actually do something, without needing to solve the unsolvable question of what any one person “really” is
#ai #artificialIntelligence #chatgpt #technology #writing -
‘A friend I can trust’: How #Americans described their relationship with #AI
More than a quarter of #US adults turn to #chatbots for personal or emotional discussions
Millions of Americans share personal info and discuss private thoughts with AI as they integrate apps like #ChatGPT intimately into their lives. Many talk to AI as they might a friend or #therapist, turning to them for emotional support and incorporating advice into their own lives
https://www.washingtonpost.com/technology/interactive/2026/09/02/27-us-adults-turn-ai-personal-emotional-social-queries/
https://archive.ph/D1BxW -
ChatGPT Sites ganha edição colaborativa e duplica velocidade de publicação. Com a nova funcionalidade "Build Together", a OpenAI reduziu para metade o tempo necessário para criar e implementar páginas web na plataforma.
-
How To Expand Images Scenarios Using ChatGPT? 🤖🖼 https://www.youtube.com/watch?v=MflKTCAy_bg 🎨 #Expand #Images #Scenarios #ChatGPT #AI
-
Habe gerade #ChatGPT was gefragt, und jede Antwort enthält jetzt eine Werbeeinblendung. *Gefühlt* sind die Antworten der #KI in dem Zuge auch gleich weniger zielführend geworden. Kann es etwa sein dass die KI darauf getrimmt wurde, Leute möglichst lange in der App zu halten?
Kommt mir bestimmt nur so vor...
-
[Veille 📣] Économie. Opacité de la dette, montages circulaires... pourquoi la bulle de l’IA rend tout le monde nerveux
https://www.ledauphine.com/economie/2026/09/12/opacite-de-la-dette-montages-circulaires-pourquoi-la-bulle-de-l-ia-rend-tout-le-monde-nerveux> Le cabinet SemiAnalysis a chiffré qu’un abonnement ChatGPT Pro à 200 dollars pouvait générer jusqu’à 14 000 dollars de coûts de calcul.
> L'existence d’une bulle ne fait aucun doute. La vraie question, c’est de savoir si cette bulle va éclater, si elle va se dégonfler, et quels seront les dommages que ça pourrait causer.
Bah tout va bien, continuons à en mettre partout.
Lien vers cette entrée dans ma veille : https://liens.declosure.ch/shaare/NyyXIA
Toute ma veille est disponible ici : https://liens.declosure.ch/
#IA #ChatGPT #économie #spéculation
#links #liens #veille
