Eateries will never be the same.

Technology giant Apple Inc. (NASDAQ: AAPL) has bought a small Canadian start-up company to help it improve machine-learning and artificial intelligence (AI).

Apple has purchased Waterloo, Ontario-based company Inductiv Inc., adding to more than a dozen AI-related acquisitions in recent years. Inductiv develops technology that uses AI to automate the task of identifying and correcting errors in data. Having clean data is important for machine learning, a popular and powerful type of artificial intelligence that helps software improve with less human intervention.

The engineering team from Inductiv joined Apple in recent weeks to work on Siri, machine learning and data science. The Inductiv acquisition is part of Apple’s broader machine-learning strategy. Apple has been upgrading the underlying technology that goes into the Siri digital assistant and other AI-powered products.

Dozens of journalists have been sacked after Microsoft decided to replace them with artificial intelligence software.

Staff who maintain the news homepages on Microsoft’s MSN website and its Edge browser – used by millions of Britons every day – have been told that they will be no longer be required because robots can now do their jobs.

Around 27 individuals employed by PA Media – formerly the Press Association – were told on Thursday that they would lose their jobs in a month’s time after Microsoft decided to stop employing humans to select, edit and curate news articles on its homepages.

A team of researchers from OpenAI recently published a paper describing GPT-3, a deep-learning model for natural-language with 175 billion parameters, 100x more than the previous version, GPT-2. The model is pre-trained on nearly half a trillion words and achieves state-of-the-art performance on several NLP benchmarks without fine-tuning.

In paper published on arXiv, a team of over 30 co-authors described the model and several experiments. The researchers’ goal was to produce an NLP system that performs well on a variety of tasks with little or no fine-tuning, and previous work had indicated that larger models might be the solution. To test that hypothesis, the team increased the size of their previous model, GPT-2, from 1.5 billion parameters to 175 billion. For training, the team collected several datasets, including the Common Crawl dataset and the English-language Wikipedia. The model was evaluated against several NLP benchmarks, matching state-of-the-art performance on “closed-book” question-answering tasks and setting a new record for the LAMBADA language modeling task.

OpenAI made headlines last year with GPT-2 and their decision not to release the 1.5 billion parameter version of the trained model due to “concerns about malicious applications of the technology.” GPT-2 is one of many large-scale NLP models based on the Transformer architecture. These models are pre-trained on large text corpora, such as the contents Wikipedia, using self-supervised learning. In this scenario, instead of using a dataset containing inputs paired with expected outputs, the model is given a sequence of text with words “masked” and it must learn to predict the masked words based on the surrounding context. After this pre-training, the models are then fine-tuned with a labelled benchmark dataset for a particular NLP task, such as question-answering.