may 5

week of 5/1/2023 - 5/5/2023

Welcome to the first weekly update from the Handy AI newsletter. Every Friday we’ll round up the week’s most interesting AI advancements, providing you with the information you need to stay in the loop.

what to know for now

  • 💬 ElevenLabs now lets users replicate a voice in numerous languages. The new Multilingual product allows users to replicate a voice in any language, maintaining a voice’s unique characteristics with extreme accuracy across seven languages.
  • 👩🏼‍💼 IBM blames AI for shift in hiring strategy. CEO Arvind Krishna discussed that around 7,800 jobs would be deprioritized in the name of AI automation going forward.
  • 🧑🏼‍💻 OpenAI expands access to the Code Interpreter model. More ChatGPT Plus subscribers can access the plugin, enabling the AI to run Python code and support file uploads/downloads for significant improvement to its data analysis capabilities.
  • 🗨️ Microsoft’s Bing Chat gets a big update. The update improves its base design, expands access to everyone, and upgrades performance on certain requests.
  • 🧩 Slack’s AI integrations are formally launched. SlackGPT is detailed and made available for some.
  • 📔 The education sector begins feeling the heat from AI. Shares suffering large drops across various education companies, including Chegg and Duolingo.

what to know for later

  • ⚖️ AI regulations are coming. The UK, US, and China begin taking AI regulation planning more seriously.
  • 📦 Amazon plans to incorporate LLM technology into Alexa. The incorporation of LLMs will give Alexa users a ChatGPT-like experience, enabling more natural conversations and better responses.
  • ❗A prominent researcher quits Google to warn the public of AI problems. Geoffrey Hinton leaves Google to free up his ability to speak on the dangers of AI.
  • 👐🏼 Open source may be the path forward. A unnamed Google researcher warns that closed-sourced AI models like OpenAI and Microsoft’s will be surpassed by the open source models the AI community is driving.
  • 🪖 Palantir launches its first AI product. The Palantir AIP platform is focused on using LLM models to automate certain parts of military operations.
  • 🧠 AI’s mind reading abilities are becoming a reality. A small test begins probing AI’s ability to read minds through “translat[ing] brain activity into a continuous stream of text”

📬 Want weekly updates delivered to your inbox? Subscribe to the Handy AI newsletter here.