SI explained: news and plain-English explainers on Super Intelligence
Plain-English explainers on SI, the name the US government adopted for artificial intelligence in September 2026, and on the ideas around it: what changed with the renaming, what did not, and the everyday concepts that come with generative SI, such as hallucinations and deepfakes. We stick to what the sources say, separate facts from commentary, and update articles when the situation moves. For the definition itself, the history of the word and how SI differs from superintelligence, read our reference guide What is SI?
Generative SI (generative AI) is the kind of SI that makes new text, images and sound from a prompt. How it learns, how text and image generators work, how many people use it, its risks, and why its output is hard to spot.
Google says its AI Overviews cannot be turned off. You can still hide these SI (AI) summaries with the official Web filter or make a link-only search your default. What works, what is only a workaround, and what each method costs you.
No single person invented SI (AI). John McCarthy named the field in 1955, Turing asked whether machines can think, Newell, Simon and Shaw wrote the first reasoning program, and neural network pioneers built today's models. Who did what, from the original documents.
Alan Turing's imitation game asks whether a judge can tell a machine from a person in a text chat. What he proposed and predicted, the claimed passes, the 2025 study where GPT-4.5 was picked as the human 73% of the time, and why passing does not prove a machine thinks.
AI companion apps are chatbots built to be friends, confidants or partners. What they are, how they work, what research shows about loneliness and manipulation, what they do with your data and how regulators are responding.
"AI psychosis" is not a diagnosis, but psychiatrists have now published cases where a chatbot reinforced delusions. What the case reports, a chat log study and OpenAI's figures actually show, and what remains unknown.
ChatGPT is a transformer trained to predict the next token, then shaped by human feedback into an assistant. We follow the whole chain, from training data to the answer on your screen, and note what OpenAI does not publish.
A token is the chunk of text an SI language model actually reads and writes. We show real tokens, explain byte pair encoding, test the 4-characters rule and explain why tokens decide context limits and prices.
A neural network is layers of simple calculations whose weights are learned from examples. We explain the parts, how training works, where the idea came from and what it does not mean.
Headlines about SI (AI) and jobs quote exposure indices, employer surveys and payroll data as if they were the same thing. We explain what each study counts, what has changed so far and who is most exposed.
Prompt injection is an attack in which someone else's text is read by an SI (AI) model as an instruction. We explain where the name comes from, why it works, two documented cases and how to reduce the risk.
One estimate says a bottle of water per conversation, another says five drops per prompt. We put the SI (AI) water numbers side by side and explain what each one measures.