Lifeboat Foundation

Deepgram has made a name for itself as one of the go-to startups for voice recognition. Today, the well-funded company announced the launch of Aura, its new real-time text-to-speech API. Aura combines highly realistic voice models with a low-latency API to allow developers to build real-time, conversational AI agents. Backed by large language models (LLMs), these agents can then stand in for customer service agents in call centers and other customer-facing situations.

As Deepgram co-founder and CEO Scott Stephenson told me, it’s long been possible to get access to great voice models, but those were expensive and took a long time to compute. Meanwhile, low latency models tend to sound robotic. Deepgram’s Aura combines human-like voice models that render extremely fast (typically in well under half a second) and, as Stephenson noted repeatedly, does so at a low price.

The advisory followed a mini-scandal over Gemini and the Indian government last month. The AI tool, responding to a query about whether Indian Prime Minister Narendra Modi was a fascist, responded that Modi had been accused of implementing policies that some had characterized as fascist. India’s deputy IT minister Rajeev Chandrasekhar called it “direct violations” of the IT Rules, 2021.

Last month, Google also suspended Gemini’s ability to generate people’s images after it showed historical inaccuracies. The company then said it would soon re-release an improved version to address the issues.

It is unclear whether Google will unblock Gemini for answering election-related queries after the elections end later this year. We have contacted Google about this and will update this story when the company responds. We are also awaiting a full list of countries where the update is live now, and we will update the post as we learn more.

This article discusses the risks of ChatGPT and Generative AI innovations are having on sustainability goals and highlights the energy and water consumption risks.

Jointly developed by investigators of the Vall d’Hebron Institute of Oncology’s (VHIO) Radiomics Group and the Bellvitge University Hospital’s Neuroradiology Unit, the Diagnosis in Susceptibility Contrast Enhancing Regions for Neuroncology (DISCERN) is an open-access deep learning tool based on the training of patterns using artificial intelligence models from information of standard magnetic resonance imaging (MRI).

Models Large Language Model Agents and swarms thereof as computational graphs reflecting the hierarchical nature of intelligence.

Graph optimization automatically improves nodes and edges.

Continue reading “Metauto-ai/GPTSwarm: 🐝 GPTSwarm: LLM agents as Graphs” »

How LLM #AI can make a patient-friendly— more understandable, more concise— hospital discharge summary for patients.

Generative artificial intelligence to transform inpatient discharge summaries to patient-friendly language and format.

This cross-sectional study, as part of a larger project to improve care delivery in our health system, was deemed exempt from institutional review board review based on the NYU Langone Health self-certification protocol. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Continue reading “Generative AI for Patient-Friendly Language in Discharge Summaries” »

Five aerospace companies are bidding for the $6 billion contract to produce new unmanned aircraft for the Air Force. The planes will be AI-piloted and will be able to perform dangerous maneuvers.

Thiruvananthapuram, Kerala: A school in Kerala is taking what may be called a revolutionary step towards revamping education with the introduction of Iris, claimed to be the first-ever AI teacher robot in the state.

The KTCT Higher Secondary School, a venture of the Kaduvayil Thangal Charitable Trust, unveiled Iris last month in collaboration with Makerlabs Edutech Private Limited. The Iris robot is designed to be more than just a robot. Built as part of the Atal Tinkering Lab (ATL) project by NITI Aayog, Iris is equipped to answer complex questions across various subjects in three different languages. It can also provide personalized voice assistance and facilitate interactive learning experiences.

From creating images, generating text, and enabling self-driving cars, the potential uses of artificial intelligence (AI) are vast and transformative. However, all this capability comes at a very high energy cost. For instance, estimates indicate that training OPEN AI’s popular GPT-3 model consumed over 1,287 MWh, enough to supply an average U.S. household for 120 years.