Applications
Where AI is actually being used: products, industries, and the work it changes.
-
Apple A14 Bionic
Apple doubled the Neural Engine to 16 cores and quoted 11 trillion operations per second, without ever naming a precision.
-
Apple A15 Bionic
Roughly 44 per cent more stated operations per second at the same, a clock and efficiency gain rather than an architectural one.
-
Fine-tuning
Continuing to train a model on your own examples to change how it behaves, not what it knows.
-
LoRA
Fine-tuning by training a small number of added parameters instead of updating the whole model.
-
MCP
An open standard letting any compatible AI client use any compatible tool server, replacing bespoke integrations.
-
Task horizon
How long a task, measured in human working time, a model can complete at a given success rate.
-
Tool use
Letting a model request an action (a search, a calculation, a database query) that your code then executes.
-
Agent harness
The code around a model that defines its tools, enforces limits, manages context and handles failure.
-
Prompt injection
Text placed where a model will read it, written to redirect the model’s behaviour.
-
Grounding
Supplying the facts in the prompt so the model answers from source material rather than memory.