Hi {{ subscriber.first_name | default: "there" }},

AI is learning to see, hear, and speak — and it needs humans for all three.

This week's list is heavy on audio, speech, visual reasoning, and video. That is not a coincidence. The next wave of AI products is multimodal, and the companies building them are hiring people who can evaluate what models see, hear, and say — not just what they write.

If you have ears, eyes, and judgment, you are more qualified than you think.

AllJobs.work now has 70,000+ remote jobs, updated daily.

Here are the roles that stood out most this week:

🎙️ Audio & Speech — AI's Next Big Push

Best QA role this week
QA Specialist - Audio Annotation & Diarization (US English) — Turing
📍 Remote (Worldwide)
Why It's Interesting: You are the final checkpoint. Contributors record real group conversations, run them through transcription and diarization, then you review the whole pipeline — audio quality, speaker labels, transcript accuracy, everything. This is not entry-level annotation. It is structured quality control on speech data, and that skill set is becoming more valuable as voice AI products ship faster.
Apply →

Accent work, real pay
English (India) Recording Specialist — micro1
📍 Remote (Worldwide) | 💰 $30–$40/hr
Why It's Interesting: AI speech models still struggle with accents, and companies are willing to pay well to fix that. This is not about reading scripts — it is about providing authentic, high-quality voice data that helps models understand how real people actually speak. If you have Indian English fluency, your voice is literally the training data.
Apply →

xAI + rare language
AI Tutor - Swahili — xAI
📍 Remote (Worldwide) | 💰 $35–$45/hr
Why It's Interesting: This is xAI (Elon Musk's AI lab) investing in multilingual voice — and Swahili is exactly the kind of language where good data is scarce and therefore more valuable. You are curating and annotating audio data to help Grok handle speech recognition, voice interactions, and cultural nuance across languages most AI systems still ignore.
Apply →

Strong pay, low barrier
AI Speech Evaluator (US College Student Required) — Marcs Hub
📍 Remote (US) | 💰 $45/hr
Why It's Interesting: $45/hr to read and record a word list with an American accent. The task sounds simple, but it is part of a growing category: accent-focused speech datasets that help AI systems understand regional pronunciation. If you are a US college student, this is one of the easiest high-paying AI gigs available right now.
Apply →

👁️ Visual, Video & Task Design

Spatial reasoning niche
Image Annotation Expert — Mercor
📍 Remote (Worldwide) | 💰 $30/hr
Why It's Interesting: This is not basic image labeling. You are answering precise visual reasoning questions about real images and diagrams — spatial relationships, structural details, edge cases. That puts you directly in the pipeline for AI systems that need to understand what they see, not just detect objects. Visual reasoning is one of the hardest open problems in AI right now.
Apply →

Wide pay range, real upside
Video Related Professional for AI Training — micro1
📍 Remote (Worldwide) | 💰 $20–$70/hr
Why It's Interesting: The $20-to-$70 range tells you something: they are hiring across experience levels, and video expertise pays more than generic annotation. If you have a background in video production, editing, cinematography, or media analysis, this is where that knowledge becomes AI-relevant. Video understanding is the frontier most people are not paying attention to yet.
Apply →

Lowest barrier in this list
Vision Data Intelligence Specialist — Toloka Annotators
📍 Remote (Worldwide) | 💰 $5/hr
Why It's Interesting: The pay is the lowest here, but the concept is worth knowing about. You record point-of-view video of everyday household tasks — dishes, vacuuming, laundry — to help train AI systems on real-world physical understanding. It is not career-defining work, but it is an interesting glimpse at where embodied AI is heading and one of the few roles that literally anyone can do.
Apply →

Build the tests AI takes
Tech Innovators: AI Task Creation — Terac
📍 Remote (US) | 💰 $28–$40/hr
Why It's Interesting: Most AI jobs ask you to evaluate answers. This one asks you to design the questions. You are creating technical scenarios, building expected responses, and assessing whether evaluation tasks are rigorous enough. That is a level above annotation — it is infrastructure work that shapes how AI systems get measured and improved.
Apply →

🌍 Language & Localization

Session coordinator role
Language Moderator: Arabic — Turing
📍 Remote (Worldwide)
Why It's Interesting: This is not typical annotation. You are coordinating multi-speaker recording sessions, managing energy and timing, and making sure conversations sound natural and authentic. Think of it as a producer role for AI speech data. If you are an Arabic speaker who can lead group dynamics and manage quality, this is a more interesting path than labeling text.
Apply →

Quirky but real
Hebrew Annotation Contributor — Welo Data
📍 Remote (Israel) | 💰 $18/hr
Why It's Interesting: You are classifying restaurant menu items from Hebrew into standardized English food categories. It sounds niche, but this is exactly how AI food platforms learn to work across languages and cultures. Your cultural knowledge is the product — and that is harder to automate than people think.
Apply →

Medical NLP + Polish
Polish Data Annotation Specialist — Docplanner Tech
📍 Remote (Poland) | 💰 63.57 PLN/hr
Why It's Interesting: This is stronger than generic annotation because you are preparing medical and linguistic datasets in Polish for NLP systems. Healthcare AI needs cleaner, more careful data than consumer products, and Polish language expertise adds a layer that makes you difficult to replace. Working directly with NLP experts also means real exposure to how production AI systems get built.
Apply →

Strong brand, solid rate
Data Annotation Specialist, Japanese — Cohere
📍 Remote (Canada) | 💰 30 CAD/hr
Why It's Interesting: Cohere is one of the top AI companies, and this role goes beyond basic labeling. You are ranking outputs, auditing accuracy across text and structured data, and recommending optimizations — that is closer to product quality work than annotation. Having Cohere on your resume also matters if you want to move deeper into AI.
Apply →

Best for translators
Machine Translation Evaluation – Japanese to English — CrowdGen
📍 Remote (Worldwide) | 💰 $18/hr
Why It's Interesting: You are not translating — you are evaluating machine translations. That distinction matters. AI translation engines improve by learning where they fail, and bilingual humans who can spot those failures are still essential. If you are a native English speaker with strong Japanese, this is steady, meaningful work.
Apply →

Low pay, real exposure
Junior Auditor – Chinese Language Expertise — Volga Partners
📍 Remote (China) | 💰 $3–$5/hr
Why It's Interesting: The pay is low, but the work is real: data labeling, annotation, and quality review focused on improving search engine results across regional Chinese languages and dialects. If you are based in China and want to build AI-related work history, this is an accessible entry point — just do not treat it as the destination.
Apply →