Claude (finally) gets a voice

Good morning, AI enthusiasts. The last major AI holdout just officially joined the voice movement, with Anthropic finally giving its assistant the flexibility to talk.

As usual with Anthropic, it’s higher late than never — and with the rollout of shiny recent models and now brand recent voice, the AI giant is shipping once more.

In today’s AI rundown:

• Anthropic’s recent Voice Mode for Claude

• Synthesia co-founder’s 3D world AI startup

• Automate project meeting documentation

• Study: AI learns reasoning through self-confidence

• 4 recent AI tools & 4 job opportunities

LATEST DEVELOPMENTS

ANTHROPIC

🗣️ Anthropic’s recent Voice Mode for Claude

The Rundown: Anthropic just announced the launch of its recent Voice mode for its Claude mobile apps, becoming one in every of the last major AI labs to enable users to have natural spoken conversations with its AI assistant.

The main points:

• The beta feature is ready to reach for English-speaking users in the approaching weeks and can run on Claude’s latest Sonnet 4 model.

• Users can flow naturally between speaking and typing, with five voice personalities available and real-time transcription displayed during chats.

• Voice mode also integrates with Google Workspace for paid subscribers, allowing Claude to access calendars, docs, and Gmail with voice commands.

• Free users receive 20-30 voice messages a month, with paid tiers getting “significantly higher” usage limits.

Why it matters: With all the foremost labs now offering voice modes, the competition shifts to execution — with features like latency, integrations, and the underlying model quality all playing a task within the user experience. The capabilities are also a jarring difference from the old-gen voices like Siri, showing how behind it truly is.

TOGETHER WITH POSTMAN

🚀 Skip the setup, ship the agent

The Rundown Postman’s Agent Generator delivers complete turnkey infrastructure with zero server setup, enabling developers to construct and deploy AI agents immediately without friction.

With Agent Generator, you may:

• Immediately spin up agent workflows

• Works with OpenAI, LangChain & more

• Test, debug, and deploy—all in Postman

Skip the setup and begin constructing today.

SPAITIAL

🌐 Synthesia co-founder’s 3D world AI startup

The Rundown: Synthesia co-founder Matthias Niessner just unveiled SpAItial, a brand new startup geared toward creating AI systems able to generating interactive 3D environments from texts and pictures.

The main points:

• The corporate is constructing Spatial Foundation Models (SFMs) that understand 3D space natively and may grasp geometry, physics, and material properties.

• SpAItial’s founding team includes former leaders from Synthesia, Google, and Meta, bringing expertise in 3D AI and neural rendering technologies.

• Early demos generated photorealistic 3D rooms from easy text prompts, with applications spanning gaming, construction, VR, and robotics.

Why it matters: While AI has mastered generating 2D images and videos, creating coherent, spatially aware 3D worlds stays a challenge. This recent breed of models could enable anyone to create complex virtual environments with just just a few words — tackling what many consider to be the subsequent frontier in AI.

AI TRAINING

📊 Automate project meeting documentation

The Rundown: On this tutorial, you’ll learn methods to create an automatic system with Zapier Agents that may turn meeting recordings into transcripts, summaries, and actionable task lists in Google Docs.

Step-by-step:

1. Visit Zapier Agents and create a “Latest Agent”

2. Configure your agent to trigger when recent audio files are uploaded to a specified folder in Google Drive

3. Add three essential tools: ChatGPT to transcribe the audio, ChatGPT again to summarize and extract motion points, and Google Docs to compile all the things right into a single document

4. Test your setup with a sample recording and activate your agent

Pro tip: In the beginning of every meeting, ask participants to obviously state their names before speaking and explicitly mention motion item assignments to assist the AI more accurately attribute tasks to team members.

PRESENTED BY ENCORD

📊 One platform for all of your AI data needs

The Rundown: Encord is a consolidated platform for multimodal AI data management, curation, and annotation, enabling teams to speed up model iteration cycles with balanced, accurately labeled datasets.

Leading AI teams use Encord’s fully customizable multimodal interface to:

• Evaluate GenAI outputs across video, audio, and text in record time

• Create VLA datasets with synchronized video, instruction, and trajectory data

• Unite PDF, image, video, audio, and DICOM labeling in a single interface

Try Encord today.

AI RESEARCH

☺️ Study: AI learns reasoning through self-confidence

The Rundown: Researchers from UC Berkeley and Yale introduced INTUITOR, an AI training method that allows language models to enhance their reasoning using internal confidence signals — eliminating the necessity for proper answers or external feedback.

The main points:

• INTUITOR measures how confident an AI feels about each word it generates, using this “gut feeling” as a guide for learning.

• As a substitute of needing correct answers to learn (like traditional AI training), the system rewards the AI when it produces responses it feels confident about.

• When tested on math problems, the strategy performed just in addition to conventional training, but showed even higher results on programming tasks.

• The AIs also began showing human-like reasoning behaviors — breaking down complex problems, planning, and explaining their pondering step-by-step.

Why it matters: Just as intuition and confidence play a big role in human learning, this study shows AI is succeeding inside the same system. This self-directed approach could possibly be especially worthwhile for tasks where there isn’t any clear “right answer” or where human expertise is proscribed, allowing AI to enterprise into unexplored knowledge areas.

QUICK HITS

🛠️ Trending AI Tools

• ⚙️ Claude Code – Anthropic’s agentic coding tool, now generally available

• 🧠 Nemotron AceReason – Nvidia’s math and code reasoning model

• 🦙 Llama-Factory – Superb-tune and train open-source LLMs with no code

• ▶️ OpusClip Thumbnail – One-click AI thumbnail generator

💼 AI Job Opportunities

• 🎧 Meta – Software Engineering Manager, Audio

• 🛠️ Palantir Technologies – Systems Engineer

• 🕴️ OpenAI – Executive Recruiter

• 🤝 Horizon3 – Partner Success Manager

📰 The whole lot else in AI today

Mistral launched Agents API for enterprise apps, introducing connectors for coding, web search, and image generation alongside memory and multi-agent orchestration.

Meta is reportedly restructuring its AI organization into two distinct teams focused on AI products and AGI foundations, aiming to speed up the corporate’s development.

Anthropic’s Claude 4 Sonnet model achieved a brand new SOTA on the ARC-AGI-2 benchmark, surpassing o3 for the highest spot on the leaderboard.

Google DeepMind teased SignGemma, an upcoming model able to translating sign language into text.

Salesforce acquired cloud data management firm Informatica for $8B, strengthening the infrastructure powering its agent-based products and platforms.

The Browser Company revealed that it’ll now not be working on its Arc browser, as a substitute fully pivoting to developing its AI-first Dia browser as a separate product.

COMMUNITY

🎥 Join our next live workshop

Join our next workshop this Friday, May thirtieth, at 4 PM EST with Dr. Alvaro Cintas, The Rundown’s AI professor. By the top of the workshop, you’ll confidently give you the chance to make use of AI coding agents to enhance your development workflow.

RSVP here. Not a member? Join The Rundown University on a 14-day free trial.

🤝 Share The Rundown, get rewards

We’ll at all times keep this text 100% free. To support our work, consider sharing The Rundown with your folks, and we’ll send you more free goodies.

See you soon,

Rowan, Joey, Zach, Alvaro, and Jason—The Rundown’s editorial team