Competency Questions (CQs) are a form of ontology functional requirements expressed as natural language questions. Inspecting CQs together with the axioms in an ontology provides critical insights into the intended scope and applicability of the ontology. CQs also underpin a number of tasks in the development of ontologies e.g. ontology reuse, ontology testing, requirement specification, and the definition of patterns that implement such requirements. Although CQs are integral to the majority of ontology engineering methodologies, the practice of publishing CQs alongside the ontological artefacts is not widely observed by the community. In this context, we present an experiment in retrofitting CQs from existing ontologies. We propose RETROFIT-CQs, a method to extract candidate CQs directly from ontologies using Generative AI. In the paper we present the pipeline that facilitates the extraction of CQs by leveraging Large Language Models (LLMs) and we discuss its application to a number of existing ontologies.

The Ai Pin from Humane, a much-hyped startup founded by former Apple employees, will cost $699. That'a according to The Verge, which obtained documents about the device ahead of its official launch on November 9. In addition, the Pin will reportedly have a monthly $24 subscription fee for access to T-Mobile's cellular network and large language models from OpenAI and Microsoft to power its smarts. The Ai Pin is a device that's about the size of a large business card that clips on to your clothing magnetically and acts as a personalized assistant controlled via voice and touch. Notably, it doesn't have a screen.

Samsung has joined the generative AI rat race by announcing its own model. Developed by Samsung Research, Gauss (named after mathematician Carl Friedrich Gauss) will power several on-device AI technologies and could make its public debut in the next few months. Samsung Gauss Language can handle tasks like translations and summarizing documents. The tech will be able to write emails for you too, Samsung says. Samsung Gauss Code is a coding assistant, while Samsung Gauss Image is a generative image model.

OpenAI's extremely popular ChatGPT service was down and non-functional for its 100 million weekly active users. The service went down just before 9AM ET. OpenAI has acknowledged the outage and said that it's also impacting the company's API services. However, the service was restored at around 10:50 AM ET. Instead of a working platform, ChatGPT users were greeted with a warning message that says it's "at capacity right now." OpenAI wrote in an error report that it had "identified an issue resulting in high error rates across the API and ChatGPT, and we are working on remediation."

Google is rolling out new features and improvements for Home and Nest, one of which could make it much easier for users to create complex home automations even if they don't know how to code. The company's experimental "help me script" feature leverages the power of generative AI to create home automation scripts from the natural language prompts users type in. They can, for instance, write "when the TV turns on after sunset, dim the living room lights and close the blinds" to instantly generate a script they can use. "Help me script" lives inside Google Home for the web, and it appears as a panel inside the script editor when users click on " Add new." All they have to do is write a prompt, press enter and then copy-paste the script results into the script editor.

Inference optimizations are critical for improving user experience and reducing infrastructure costs and power consumption. In this article, we illustrate a form of dynamic execution known as speculative sampling to reduce the overall latency of text generation and compare it with standard autoregressive sampling. This can be used together with model-based optimizations (e.g. quantization) to provide an optimized solution. Both sampling methods make use of KV caching. A Jupyter notebook and some sample executions are provided.

Translation Quality Evaluation (TQE) is an essential step of the modern translation production process. TQE is critical in assessing both machine translation (MT) and human translation (HT) quality without reference translations. The ability to evaluate or even simply estimate the quality of translation automatically may open significant efficiency gains through process optimisation. This work examines whether the state-of-the-art large language models (LLMs) can be used for this purpose. We take OpenAI models as the best state-of-the-art technology and approach TQE as a binary classification task. On eight language pairs including English to Italian, German, French, Japanese, Dutch, Portuguese, Turkish, and Chinese, our experimental results show that fine-tuned gpt3.5 can demonstrate good performance on translation quality prediction tasks, i.e. whether the translation needs to be edited. Another finding is that simply increasing the sizes of LLMs does not lead to apparent better performances on this task by comparing the performance of three different versions of OpenAI models: curie, davinci, and gpt3.5 with 13B, 175B, and 175B parameters, respectively.

OpenAI recently announced multiple new features for ChatGPT and other artificial intelligence tools during its recent developer conference. The upcoming launch of a creator tool for chatbots, called GPTs (short for generative pretrained transformers), and a new model for ChatGPT, called GPT-4 Turbo, are two of the most important announcements from the company's event. This isn't the first time OpenAI has given ChatGPT a new model. Earlier this year, OpenAI updated the algorithm for ChatGPT from GPT-3.5 to GPT-4. Are you curious how the GPT-4 Turbo version of the chatbot will be different when it rolls out later this year?

This article is from Big Technology, a newsletter by Alex Kantrowitz. It's been a year of glossy profiles, breathless accolades, and billions in new funding for OpenAI, but the ChatGPT maker is far more vulnerable than the popular narrative suggests. Amid a seemingly unstoppable ascent, the company is facing fierce competition, a rising open-source movement, and pressure to deliver hits in an unpredictable discipline. While its marquee product has become practically synonymous with A.I., its perch atop the field is less than rock solid. OpenAI's weakness stems in part from its strength. It popularized generative A.I. by taking others' innovations--like the transformer model--and building stellar products on top of them.

Xbox has teamed up with a startup called Inworld AI to create a generative AI toolset that developers can use to create games. It's a multi-year collaboration, which the Microsoft-owned brand says can "assist and empower creators in dialogue, story and quest design." Specifically, the partners are looking to develop an "AI design copilot" that can turn prompts into detailed scripts, dialogue trees, quests and other game elements in the same way people can type ideas into generative AI chatbots and get detailed scripts in return. They're also going to work on an "AI character runtime engine" that developers can plug into their actual games, allowing players to generate new stories, quests and dialogues as they go. On Inworld's website, it says its technology can "craft characters with distinct personalities and contextual awareness that stay in-world."