AI News Today

Today's podcast discusses Cognition AI's new tool Devin, which takes a leap in AI-powered coding by autonomously creating entire engineering codes. Also, we report on AI image generating service, Midjourney's feature rollout that allows the consistent recreation of characters across new images, solving a major challenge in the industry. Concluding, we explore Deepgram's real-time text-to-speech API called Aura that enables developers to create AI agents simulating customer service reps, benefiting businesses seeking cost-effective AI solutions.

Cohere introduces their groundbreaking Command-R model, seemingly superior to OpenAI's GPT-3.5 Turbo, making waves in the AI for business scene. Apple experiments with AI-powered ads within their App Store, potentially shifting the balance in the ad tech world. Finally, French startup Nijta emerges with AI speech anonymization technology, promising a privacy-centered solution to industries handling sensitive voice data.

On today's episode, we discuss Elon Musk's plan to open-source xAI's AI chatbot, Grok, amidst a lawsuit against OpenAI. Next, we explore the outcry at the UK's Alan Turing Institute over diversity concerns in leadership roles. Finally, we dive into Pika Labs' innovative sound effects addition to their user's videos. Tune in for these stories and more.

This episode discusses the latest AI news. We examine how large language models are outperforming humans in predicting neuroscience experiment success rates with an 81.4% accuracy. Next, explore the launch of SauLM-7B, an open-source, text-generating, AI model specifically designed for legal work in collaboration with Equall.ai and various universities. Lastly, we anticipate the release of a new neural voice tool by Microsoft, crafted to aid individuals with speech disabilities. Tune in for these insightful discussions.

In this episode, we discuss OpenAI welcoming back former CEO Sam Altman to its board of directors, together with three renowned new members, in a move to address diversity and past relationships. We explore Google's latest feature for its Gemini AI chatbot, a tool granting users the power to modify the bot's responses for better precision and control. Lastly, we delve into an intriguing partnership between Soul Machines and Authentic Brands Group, delivering a realistic AI-powered digital version of Marilyn Monroe that's capable of real-time interaction.

Inflection introduces Inflection-2.5, their personal AI that matches global models like GPT-4 and Gemini, with enhanced coding, mathematics and real-time web search capabilities. Meanwhile, Adobe is taking mobile content creation to the next level with their latest Adobe Express app, integrated with Firefly AI models for swift and efficient content generation and editing. Finally, Reddit plans to utilize AI in developing a harassment filter to assist volunteer moderators in dealing with harmful content, amidst ongoing user concerns and their impending IPO.

Meta is investing in a single AI system to power video recommendations across platforms, shifting from its previous approach of individual models. Additionally, Writable, an AI tool recently acquired by Houghton Mifflin Harcourt, is aiding teachers in grading essays by using ChatGPT to provide feedback. Lastly, Zapier acquired the team from Vowel and subsequently launched Zapier Central for AI bot creation while appointing Vowel's CEO as Director of AI.

This episode recaps a current legal battle as Elon Musk accuses OpenAI, the AI lab he left in 2018, of pivoting from their nonprofit status to prioritize profit, sparked by their multi-billion -dollar partnership with Microsoft. We also cover Deepmind alumni's launch of Haiper, a revolutionary AI-powered video generation tool developed with millions in seed and pre-seed funding. We wrap up the episode by discussing Cloudflare's development of a Firewall for AI, a protection layer specially crafted for Large Language Models to mitigate potential abuses.

OpenAI expands the functionality of ChatGPT with a Read Aloud feature that translates text into speech in multiple languages. Meanwhile, Huawei develops the Ascend 910B AI chip, rivaling the performance of Nvidia's A100 GPU - a significant advance in the AI chips industry. In entertainment, Madonna partners with startup Runway to create AI-generated concert visuals, showcasing the unlimited potential and equally challenging ethical considerations of AI.

In this episode, we discuss the launch of Anthropic's groundbreaking suite of AI models, Claude 3, which outperforms rival products from OpenAI and Google. We also touch on Google's adoption of AI across its platforms, including the 'Photomath' application, which now uses AI to provide step-by-step solutions to mathematical problems. Wrapping up, we delve into India's new advisory on AI, requiring tech companies to seek government approval for new models, signaling a more regulated future for the AI industry in the country.