Memory Leak — #18
VC Astasia Myers’ perspectives on machine learning, cloud infrastructure, developer tools, open source, and security. Sign up here.
Why does this matter? PromptLayer argues that prompts are the new gold, and the most important part of any LLM app. With the rise of prompt engineering we’ve seen a handful of LLM optimization startups emerge like PromptLayer that help with prompt management, A/B testing, fine-tuning, and monitoring.
Arcwise AI — GPT Copilot for Sheets
Arcwise AI let’s you use the AI behind ChatGPT to understand, clean, and ingest data in Google Sheets with text commands! Just press Cmd+Shift+1 / Ctrl+Shift+1. It also allows users to ask what a spreadsheet does and what formula to use.
Why does this matter? While we are seeing how foundational models can disrupt creative fields like content creation, they could have the biggest impact on manual, mundane tasks like data prep and cleaning. In Anaconda’s State of Data Science 2022 survey of 3,493 individuals, respondents indicated they spend about 37.75% of their time on data preparation and cleansing. It remains the largest area that data scientists spend their time.
Raycast AI: The Magic of AI, Right on Your Mac
Write smarter, code faster and answer questions quicker with ChatGPT in Raycast — just one keystroke away. Leverage Raycast’s integration with macOS to enjoy an entirely new layer of context that makes using AI across all your applications a breeze including creating AI commands to take your workflows. It can rewrite text and generate fresh ideas. Just describe your task and let Raycast do the rest.
Why does this matter? Raycast AI is an example outside of Github Copilot of how developer tools are being augmented with AI.
Buzzy Storytelling Startup Tome Raises $43 Million From A Who’s Who In AI
Now valued at $300 million while still pre-revenue, per a source, Tome says it’s the fastest productivity software maker to ever reach 1 million users since its September 2022 release.
Why does this matter? It’s a race between the GenAI-native SaaS products and the incumbents who are retrofitting their products with AI capabilities. Tome takes this battle to Google Slides and Microsoft PowerPoint. Tome allows users to type in a prompt that generates entire narratives from scratch or creates additional content pages. It will be interesting to see if and how big tech responds.
Spotify Launches ‘DJ,’ a New Feature Offering Personalized Music With Ai-Powered Commentary
Spotify launched a new AI feature called “DJ” to better personalize the music listening experience for its users. Similar to a radio DJ, Spotify’s DJ feature will deliver a curated selection of music alongside AI-powered spoken commentary about the tracks and artists you like, using what Spotify says is a “stunningly realistic voice.”
Why does this matter? The application of generative or synthetic audio is more nascent than image, video, or text. Lyrebird acquired by Descript led the way by helping creators easily adjust their commentary in post-processing. Now we are seeing the rise of synthetic voice startups like Coqui, Rime.ai, Murf.ai, Resemble.ai, Uberduck.ai, Synthesys.io, Elevenlabs.io, among others. Spotify’s AI DJ is an interesting application of AI voice technologies.
My Postgres Landscape
Why does this matter? The first Postgres release was in January 1997 and was originally used for Online Transactional Processing (OLTAP). Over the past 26 years, it’s become a hardened and popular transactional database. Recently a handful of startups have emerged around Postgres like Supabase and Neon. Many databases are now Postgres compatible like Cockroach. Even more interesting are the startups extending Postgres to analytical use cases like Hydra and ML like PostgresML. There are tradeoffs for using Postgres for everything, but it is clear people love Postgres and it’s here to stay.
⭐️Claypot — Founding Engineer (Infra)
⭐️Grit — Design Engineer
⭐️Speakeasy — Founding UX Lead