Not every performer began their career as a child star -- but thanks to artificial intelligence (AI) we can now all catch a glimpse of what their childhood fame might have looked like. It comes after an image of an infant Elon Musk -- showing the billionaire as a toddler wearing brown overalls -- gave the internet baby fever. The adorable (or possibly disturbing?) photo is not from the billionaire's own family archives, however. Like these celebrities, it too was made with the help of artificial intelligence. The AI-made toddler version of Ed Sheeran has his soulful blue eye and his vibrant red hair.

Voice assistants help users make phone calls, send messages, create events, navigate, and do a lot more. However, assistants have limited capacity to understand their users' context. In this work, we aim to take a step in this direction. Our work dives into a new experience for users to refer to phone numbers, addresses, email addresses, URLs, and dates on their phone screens. Our focus lies in reference understanding, which becomes particularly interesting when multiple similar texts are present on screen, similar to visual grounding. We collect a dataset and propose a lightweight general-purpose model for this novel experience. Due to the high cost of consuming pixels directly, our system is designed to rely on the extracted text from the UI. Our model is modular, thus offering flexibility, improved interpretability, and efficient runtime memory utilization.

Concepts benefit natural language understanding but are far from complete in existing knowledge graphs (KGs). Recently, pre-trained language models (PLMs) have been widely used in text-based concept extraction (CE). However, PLMs tend to mine the co-occurrence associations from massive corpus as pre-trained knowledge rather than the real causal effect between tokens. As a result, the pre-trained knowledge confounds PLMs to extract biased concepts based on spurious co-occurrence correlations, inevitably resulting in low precision. In this paper, through the lens of a Structural Causal Model (SCM), we propose equipping the PLM-based extractor with a knowledge-guided prompt as an intervention to alleviate concept bias. The prompt adopts the topic of the given entity from the existing knowledge in KGs to mitigate the spurious co-occurrence correlations between entities and biased concepts. Our extensive experiments on representative multilingual KG datasets justify that our proposed prompt can effectively alleviate concept bias and improve the performance of PLM-based CE models.The code has been released on https://github.com/siyuyuan/KPCE.

There are so many good smart displays out there that it can be hard to choose which one to buy. Right now, one of our favorites, Amazon's second-generation Echo Show 8, is running down to $75 from $130 -- a 42 percent discount and just $5 off its lowest price. There are other available options on sale, like an adjustable stand or Blink Mini, but expect to pay a little extra for those. The Echo Show 8 is part speaker, part tablet, with TV shows and movies available from streamers like Netflix, Hulu and, of course, Prime Video. These come alongside music from Spotify, Apple Music and Amazon Music.

Justine Bateman told Fox News Digital using artificial intelligence to write a script is not solving any problems because there is no lack of talent in the industry. With big bank accounts, celebrities have begun to invest money in companies using artificial intelligence, predominately startups. Within the past few years, "The Wolf of Wall Street" actor Leonardo DiCaprio and "Iron Man" himself, Robert Downey Jr., have both reportedly invested millions, along with their respective venture capital firms, into AI companies designed to impact the environment. Other stars, including Ashton Kutcher and Black Eyed Peas singer and rapper will.i.am., are also exploring the world of AI, something Kutcher believes is deeply intertwined into a successful future. Ashton Kutcher believes AI is the future and a good thing for humanity.

SmartNews' Head of Global Trust and Safety is calling for new regulation on artificial intelligence (AI) to prioritize user transparency and ensure human oversight remains a crucial component for news and social media recommender systems. "We need to have guardrails," Arjun Narayan said. "Without humans thinking through everything that could go wrong, like bias creeping into the models or large language models falling into the wrong hands, there can be very serious consequences for humanity." Narayan, who previously worked on Trust and Safety for Google and Bytedance, the company behind TikTok, said it is essential for companies to recognize opt-in and opt-outs when using large language models (LLMs). As a default, anything being fed to an LLM will be assumed training data and collected by the model.

Voice digital assistants must keep up with trending search queries. We rely on a speech recognition model using contextual biasing with a rapidly updated set of entities, instead of frequent model retraining, to keep up with trends. There are several challenges with this approach: (1) the entity set must be frequently reconstructed, (2) the entity set is of limited size due to latency and accuracy trade-offs, and (3) finding the true entity distribution for biasing is complicated by ASR misrecognition. We address these challenges and define an entity set by modeling customers true requested entity distribution from ASR output in production using record deduplication, a technique from the field of entity resolution. Record deduplication resolves or deduplicates coreferences, including misrecognitions, of the same latent entity. Our method successfully retrieves 95% of misrecognized entities and when used for contextual biasing shows an estimated 5% relative word error rate reduction.

Large language models (LLMs) have become mainstream technology with their versatile use cases and impressive performance. Despite the countless out-of-the-box applications, LLMs are still not reliable. A lot of work is being done to improve the factual accuracy, consistency, and ethical standards of these models through fine-tuning, prompting, and Reinforcement Learning with Human Feedback (RLHF), but no systematic analysis of the responses of these models to different categories of statements, or on their potential vulnerabilities to simple prompting changes is available. In this work, we analyze what confuses GPT-3: how the model responds to certain sensitive topics and what effects the prompt wording has on the model response. We find that GPT-3 correctly disagrees with obvious Conspiracies and Stereotypes but makes mistakes with common Misconceptions and Controversies. The model responses are inconsistent across prompts and settings, highlighting GPT-3's unreliability. Dataset and code of our analysis is available in https://github.com/tanny411/GPT3-Reliability-Check.

OpenAI has released the Chat Generative Pre-trained Transformer (ChatGPT) and revolutionized the approach in artificial intelligence to human-model interaction. Several publications on ChatGPT evaluation test its effectiveness on well-known natural language processing (NLP) tasks. However, the existing studies are mostly non-automated and tested on a very limited scale. In this work, we examined ChatGPT's capabilities on 25 diverse analytical NLP tasks, most of them subjective even to humans, such as sentiment analysis, emotion recognition, offensiveness, and stance detection. In contrast, the other tasks require more objective reasoning like word sense disambiguation, linguistic acceptability, and question answering. We also evaluated GPT-4 model on five selected subsets of NLP tasks. We automated ChatGPT and GPT-4 prompting process and analyzed more than 49k responses. Our comparison of its results with available State-of-the-Art (SOTA) solutions showed that the average loss in quality of the ChatGPT model was about 25% for zero-shot and few-shot evaluation. For GPT-4 model, a loss for semantic tasks is significantly lower than for ChatGPT. We showed that the more difficult the task (lower SOTA performance), the higher the ChatGPT loss. It especially refers to pragmatic NLP problems like emotion recognition. We also tested the ability to personalize ChatGPT responses for selected subjective tasks via Random Contextual Few-Shot Personalization, and we obtained significantly better user-based predictions. Additional qualitative analysis revealed a ChatGPT bias, most likely due to the rules imposed on human trainers by OpenAI. Our results provide the basis for a fundamental discussion of whether the high quality of recent predictive NLP models can indicate a tool's usefulness to society and how the learning and validation procedures for such systems should be established.

This project presents a deep learning approach to generate monophonic melodies based on input beats, allowing even amateurs to create their own music compositions. Three effective methods - LSTM with Full Attention, LSTM with Local Attention, and Transformer with Relative Position Representation - are proposed for this novel task, providing great variation, harmony, and structure in the generated music. This project allows anyone to compose their own music by tapping their keyboards or ``recoloring'' beat sequences from existing works.