Nolege News

Category

Computer Science

All Nolege News articles on Computer Science.

7 articles

You can patch a bug in code. You cannot patch a model that learned the wrong thing

Computer Science

You can patch a bug in code. You cannot patch a model that learned the wrong thing

5 Sept 2026 · 8 min read

A software vulnerability lives in a line someone wrote and can be rewritten. An AI vulnerability lives in a learned function nobody wrote, and the most important one — prompt injection — has no accepted fix at all.

Read more →
Training the model is the one-time cost. Answering your question is the bill that never stops

Computer Science

Training the model is the one-time cost. Answering your question is the bill that never stops

3 Sept 2026 · 9 min read

Everyone quotes the electricity a model burned during training. For anything widely used, that is the smaller number. Here is what a GPU is actually doing, why memory rather than arithmetic is usually the limit, and why published per-query energy figures differ by a factor of a hundred.

Read more →
Nobody has to break UPI. They only have to get you to press 'Pay'.

Computer Science

Nobody has to break UPI. They only have to get you to press 'Pay'.

2 Sept 2026 · 9 min read

Almost no digital payment fraud in India involves cracking encryption. The overwhelming majority are transactions the victim authorised correctly — which is why an OTP protects the bank far better than it protects you.

Read more →
Your model is 99% accurate. That is usually the first thing to be suspicious about

Computer Science

Your model is 99% accurate. That is usually the first thing to be suspicious about

31 Aug 2026 · 8 min read

A model that scores brilliantly on your test set and collapses in deployment is not unlucky. It is the normal outcome of four specific mistakes, and each one has a name, a mechanism and a fix.

Read more →
An image generator never learned to draw. It learned to remove noise — and that is the whole trick

Computer Science

An image generator never learned to draw. It learned to remove noise — and that is the whole trick

29 Aug 2026 · 8 min read

The machine that turns a sentence into a picture was trained on exactly one task: predicting the static that was added to a photograph. Everything else — the prompt, the style, the strange failures, the copyright fight — follows from that.

Read more →
Why an AI chatbot invents a fact and says it with total confidence

Computer Science

Why an AI chatbot invents a fact and says it with total confidence

26 Aug 2026 · 8 min read

A language model will give you a citation that does not exist, in the same calm tone it uses for things that are true. That is not a bug someone forgot to fix — it follows from how the machine works. Here is what is actually happening, and what reliably reduces it.

Read more →
Machine learning, explained without the jargon

Computer Science

Machine learning, explained without the jargon

25 Jul 2026 · 4 min read

Machine learning is behind spam filters, recommendations and modern AI — but the core idea is simpler than the buzzwords suggest. Here is how a machine actually 'learns'.

Read more →

Get the science, minus the hype

Clear, occasional updates on nanotech, materials and research. No spam, unsubscribe anytime.