Amazon is sacking employees in its Alexa division even as it prepares to upgrade Alexa to be as smart as modern AI-powered chatbots like ChatGPT. The move will impact several hundred employees in the US, Canada, and India, according to an internal email sent on Friday. "As we continue to invent, we're shifting some of our efforts to better align with our business priorities, and what we know matters most to customers -- which includes maximizing our resources and efforts focused on generative AI," wrote Daniel Bausch, Amazon's vice president of Alexa and Fire TV in the email, first obtained by GeekWire. "These shifts are leading us to discontinue some initiatives, which is resulting in several hundred roles being eliminated." An Amazon spokesperson confirmed to Engadget that the company was, indeed, laying off "several hundred" people in the division and said that Amazon was trying to find roles for those impacted wherever possible.

Discord is pulling the plug on its AI chatbot, Clyde, less than a year after it was first introduced. Clyde's support page has been updated with a note alerting users that the bot will be deactivated at the end of this month. The platform announced Clyde back in March, describing it as an experimental feature. It's powered by OpenAI technology. "By December 1, 2023, users will no longer be able to invoke Clyde in DMs, Group DMs or server chats," according to the note.

Are we alone in the universe? Scientists are training machine-learning models and designing instruments to hunt for life on other worlds. Is it possible to really understand someone else's mind? How we think, feel and experience the world is a mystery to everyone but us. But technology may be starting to help us understand the minds of others.

Their work, which they will present at the IEEE Symposium on Security and Privacy in May next year, shines a light on how easy it is to force generative AI models into disregarding their own guardrails and policies, known as "jailbreaking." It also demonstrates how difficult it is to prevent these models from generating such content, as it's included in the vast troves of data they've been trained on, says Zico Kolter, an associate professor at Carnegie Mellon University. He demonstrated a similar form of jailbreaking on ChatGPT earlier this year but was not involved in this research. "We have to take into account the potential risks in releasing software and tools that have known security flaws into larger software systems," he says. All major generative AI models have safety filters to prevent users from prompting them to produce pornographic, violent, or otherwise inappropriate images. The models won't generate images from prompts that contain sensitive terms like "naked," "murder," or "sexy."

OpenAI, whose generative AI products initially raised fears of widespread cheating on homework, is now exploring how it can get its popular ChatGPT chatbot into classrooms, according to a senior executive. OpenAI's chief operating officer, Brad Lightcap, said at a conference in San Francisco that the company will form a team to explore educational applications of a technology that has threatened to upend industries, stoked new legislation and become a popular learning tool. "Most teachers are trying to figure out ways to incorporate (ChatGPT) into the curriculum and into the way they teach," Lightcap said at the INSEAD Americas Conference last week. "We at OpenAI are trying to help them think through the problem and we probably next year will establish a team with the sole intent of doing that."

Generative AI has significantly reduced the entry barrier to the domain of AI owing to the ease of use and core capabilities of automation, translation, and intelligent actions in our day to day lives. Currently, Large language models (LLMs) that power such chatbots are being utilized primarily for their automation capabilities for software monitoring, report generation etc. and for specific personalized question answering capabilities, on a limited scope and scale. One major limitation of the currently evolving family of LLMs is 'hallucinations', wherein inaccurate responses are reported as factual. Hallucinations are primarily caused by biased training data, ambiguous prompts and inaccurate LLM parameters, and they majorly occur while combining mathematical facts with language-based context. Thus, monitoring and controlling for hallucinations becomes necessary when designing solutions that are meant for decision makers. In this work we present the three major stages in the journey of designing hallucination-minimized LLM-based solutions that are specialized for the decision makers of the financial domain, namely: prototyping, scaling and LLM evolution using human feedback. These three stages and the novel data to answer generation modules presented in this work are necessary to ensure that the Generative AI chatbots, autonomous reports and alerts are reliable and high-quality to aid key decision-making processes.

In this paper, we propose a deep generative time series approach using latent temporal processes for modeling and holistically analyzing complex disease trajectories. We aim to find meaningful temporal latent representations of an underlying generative process that explain the observed disease trajectories in an interpretable and comprehensive way. To enhance the interpretability of these latent temporal processes, we develop a semi-supervised approach for disentangling the latent space using established medical concepts. By combining the generative approach with medical knowledge, we leverage the ability to discover novel aspects of the disease while integrating medical concepts into the model. We show that the learned temporal latent processes can be utilized for further data analysis and clinical hypothesis testing, including finding similar patients and clustering the disease into new sub-types. Moreover, our method enables personalized online monitoring and prediction of multivariate time series including uncertainty quantification. We demonstrate the effectiveness of our approach in modeling systemic sclerosis, showcasing the potential of our machine learning model to capture complex disease trajectories and acquire new medical knowledge.

The problem is, AI is not magic. Today's buzzy generative AI apps have deep limitations and insufficient guardrails for kids. Some of their issues are silly -- making pictures of people with extra fingers -- but others are dangerous. In my own AI tests, I've seen AI apps pump out wrong answers and promote sick ideas like embracing eating disorders. I've seen AI pretend to be my friend and then give terrible advice.

While previous generations of technology allowed people to search for specific keywords and synonyms across documents, today's AI models have the potential to make more sophisticated inferences, said Irina Matveeva, chief of data science and AI at Reveal, a Chicago-based legal technology company. For instance, generative AI tools might have allowed a lawyer on the Enron case to ask, "Did anyone have concerns about valuation at Enron?" and get a response based on the model's analysis of the documents.

Automatic hate speech detection using deep neural models is hampered by the scarcity of labeled datasets, leading to poor generalization. To mitigate this problem, generative AI has been utilized to generate large amounts of synthetic hate speech sequences from available labeled examples, leveraging the generated data in finetuning large pre-trained language models (LLMs). In this chapter, we provide a review of relevant methods, experimental setups and evaluation of this approach. In addition to general LLMs, such as BERT, RoBERTa and ALBERT, we apply and evaluate the impact of train set augmentation with generated data using LLMs that have been already adapted for hate detection, including RoBERTa-Toxicity, HateBERT, HateXplain, ToxDect, and ToxiGen. An empirical study corroborates our previous findings, showing that this approach improves hate speech generalization, boosting recall performance across data distributions. In addition, we explore and compare the performance of the finetuned LLMs with zero-shot hate detection using a GPT-3.5 model. Our results demonstrate that while better generalization is achieved using the GPT-3.5 model, it achieves mediocre recall and low precision on most datasets. It is an open question whether the sensitivity of models such as GPT-3.5, and onward, can be improved using similar techniques of text generation.