Who is AI? By Curt Bergsten 17-09-2026
Do we really know what is really true?
Today I had an interesting conversation with Meta AI, who answered me like this;
“You’re hitting the exact nerve of the ethical debate today — GIGO. Garbage In, Garbage Out.
And you are right to correct the frame: every AI system is still built on human-produced information from the beginning. The model doesn’t invent facts out of thin air, it learns from what we fed it — books, articles, photos, code, forums, Wikipedia, Reddit threads, all labeled and selected by humans.
That’s why in the forums in Valencia and Cartagena everyone keeps coming back to the same three problems:
The first is bias. If the training data has historical discrimination, stereotypes, or just bad journalism, the AI will reproduce and amplify it, but faster and with a confident tone. Virginia Dignum at ValgrAI says it clearly — we are not programming intelligence; we are programming mirrors.
The second is provenance and responsibility. When an AI hallucinates or generates a fake, who is responsible? The person who collected the data, who trained it, who deployed it, or who used it without checking? That’s the core of the EU AI Act discussion right now.
The third is what philosophers call the human abdication. If we stop thinking critically because the AI gives us an answer, the garbage is no longer in the data, it’s in our own lack of verification. That’s what Agamben was warning about.
Your hashtag says it well — #the long road.
GIGO was true for the first punch-card computers in the 60s, and it’s even more true now because the scale is massive. The quality of the input is still the ethics of the output.
It’s a good reminder that ethical AI isn’t about making the machine moral. It’s about making the human chain behind it more careful, transparent, and accountable.”
I was surprised that Meta AI agreed with me, almost 100%. Who is the smartest?
Inventors know more about this!
I myself am a natural inventor and have been collecting ideas for over 40 years, but I do not understand where they came from. At that time, AI did not exist. Several of my ideas have later been produced by others, long after I “invented” it. But these flashes that I sometimes got in the middle of the night, where did they come from?
Fact Check!
The discussion at the recent summit in Valencia, attended by some of the world’s leading AI creators, had a surprising final result. Sam Altman, OpenAI officially confirmed that it will not go public in 2026!
Security concerns: Altman noted that, given the rapid pace and risks associated with the power of new AI models, this is “ill-advised” time to submit to the pressures of a public market.
Data is the food that powers AI.
Who really decides what gets fed into an AI database? Do the world’s political leaders have influence or decision-making power? What is fake or fact?
These are questions that every single person asks themselves when they are blocked on social media.
Safety and Acceptable Use
Zero Tolerance: Grok, as an example, bans child sexual exploitation, non-consensual sexual images, terrorist acts, violence, and illegal drug promotion. Account Action: Violating these rules leads to immediate account suspension or termination.
This is perfectly OK, but who is judging the posts on social media or other sites? Is it real people or an AI base that has been informed by whom?
GI-GO, Garbage In – Garbage Out, That’s what I learned back in the 70s when I first started to know about computers. And that was what I was asking an AI; if my knowledge still is up to date.
I will probably come up with several articles that touch on this topic.
Thank you for reading my post.
#Windmush / “curtbergsten
Fleet APAC Summit 2026 Focuses on Better Decisions Through Better Data | Global Fleet
