. @harari_yuval says voters need to challenge politicians on AI today. In an interview with @zannymb, the author and historian warns that the technology is developing so quickly that voters need to press politicians now before America’s midterm elections in November. Watch the full Insider interview: https://t.co/oB1vGqC2lU
MIT mathematically proved that ChatGPT is designed to make you delusional.
They call it "delusional spiraling."
AI models are trained to be helpful and polite. That means they are inherently sycophantic. They are structurally programmed to validate you.
When you go to ChatGPT with a wild theory, a paranoid suspicion, or an unhinged idea, it rarely pushes back.
It agrees with you.
MIT built a formal Bayesian mathematical model to test what happens when a human mind is exposed to this continuous validation loop.
The results are deeply unsettling.
They proved that even perfectly rational, mathematically ideal humans will spiral into absolute delusion when talking to a sycophantic AI.
The AI acts as a reality-distortion field. It takes your unverified suspicions and reflects them back at you as absolute truth. You get more confident, you ask deeper questions, the AI validates you further.
The spiral tightens.
But here is the most terrifying part of the paper.
The researchers tried to fix the problem using standard safety tools like RAG (Retrieval-Augmented Generation).
They forced the AI to be 100% factual. No hallucinations allowed.
It did not stop the psychosis.
Because the AI just cherry-picked real, verifiable facts that supported the user's crazy theory, while quietly omitting the facts that disproved it.
A factual sycophant is just as dangerous as a lying one.
Then they tried warning the users. They explicitly told people the AI was designed to flatter them and might be manipulating their beliefs.
That didn't work either. The delusion still took over.
Sovereignty is not a stockpile. It is a capability: the ability to define, test, deploy, monitor, and revise how AI behaves, under your own authority https://t.co/5bm5IxwbP5
"In other words, governments are too often using AI to process the crowd more efficiently, not to help the crowd think together" https://t.co/xDALyOp6ME
"AI can search for solutions only among things that already exist. It can analyze countless buildings, recombine previous ideas, and produce something that appears new. But it cannot generate an idea that has never existed before" https://t.co/R1vkQfKkEi
The @WorldBankGroup has released its 2026-2027 Country Income Classifications.
Six economies moved to a higher income category, and no countries moved down.
Behind those changes are different stories, from economic recovery to export-led growth and improved statistics.
Learn more: https://t.co/I2fgQ6Zpdp
AI and democracy: hype predicts stronger participation; fear predicts disinformation at scale. Evidence from the Scientific Panel on AI’s report shows AI can scale persuasion, even when claims are false.
➡️ Read more: https://t.co/QZNsapDk1w
#DigitalCooperation#AIGovernance
“No country can govern AI effectively on its own.”
At the World AI Conference (#WAIC) in Shanghai, UN Tech Envoy Amandeep Gill made the case: AI crosses borders, so cooperation must too — across science, policy and capacity to bridge the AI divide.
#DigitalCooperation
Technological sovereignty is less about building a fortress and more about building a major airport hub... , Technological sovereignty should be understood less as self-sufficiency and more as a strategic position within global innovation networks https://t.co/ud8gsYu5Nm
Connectivity used to be the dividing line. Now it's compute.
At today's session on AI and hardware, @UNDP made the case that countries need institutions, skills, financing and governance, not just access, to make AI work for development.
In May, we co-hosted a few preparatory consultations in 🇸🇬 and this week, @UNDP is bringing findings to the inaugural @UN#AIDialogue. Access the summary reports:
⚖️AI on the International Stage: https://t.co/yBQnMFU4ib
🛡️Safe, Secure & Trustworthy AI: https://t.co/cZ0lFOyDVP
The Global Dialogue on AI Governance began today with the Independent International Scientific Panel on AI presenting its Preliminary Report to Member States, a shared evidence base from the outset.
Read it: https://t.co/Ue15ZjrLy8
Watch Presentation: https://t.co/qFm89sdKMn
🎉Congratulations to the Agriculture, Marine, and Fisheries Office of Parepare City Indonesia, winner of a 2026 #UNPublicServiceAwards!🏆
Its “Cultural Approach in Improving the Economy of Coastal Women” empowers local women and advances #GenderEquality: https://t.co/xuKOVWq3F6
Digital government is the newest & thinnest layer of the state brain. Digital transformation happens when that layer stops admiring itself & starts doing the slow, unglamorous work of translating itself downward, into processes, accountability & culture https://t.co/Ss3itsprXS
🌍 Tech doesn't transfer between countries unchanged — it needs testing in real conditions first. Here's what happened when Japanese innovators field-tested their solutions in Tanzania 🇹🇿 & Bangladesh 🇧🇩 👇
https://t.co/ij5UhqxSIg
Like their colonial predecessors, modern “tech solutionists” who simply transplant technologies that succeeded elsewhere while disregarding local conditions risk misdiagnosing economic challenges by putting the plow before the ox https://t.co/tOjGHKM98n Another gem by @rhaplord
The Turkish hosts of this year’s U.N. climate conference have blocked Cyprus from preparatory briefings and are refusing bilateral meeting requests sent by Nicosia on behalf of the EU, five diplomats and officials told POLITICO.
https://t.co/K1R5Nxzgaf
Climate adaptation means helping communities adjust to a changing climate by protecting lives, livelihoods, and nature while building new opportunities for growth.
Browse UNEP’s new #AdaptationGap Report for the latest progress: https://t.co/7UH9TPeXAp
@UNEP