SurviAGI
All markets

Market

Computing and engineering research

How far AI got on each kind of work in this market, and the update that showed it.

Work

2

2 with evidence

Furthest level reached

L2

Partial automation

Evidence

98

15 pulled back by an evidence tier

Work in this market

Open a row to read the updates behind its level.

Develop algorithms and study computational properties55 evidence12 tasksL2Partial automation

Open this work →

PublishedClaimedEvidence tierUpdateWhen
L2Partial automationL3↓ T3 allows only L2T3Vendor self-reportAMIE matches doctor diagnoses in 90% of 100 patient interactions with 0 safety stops source2026-10-08
L2Partial automationL3↓ T3 allows only L2T3Vendor self-reportGoogle's AMIE studied prospectively in real clinic, results in The Lancet source2026-10-08
L1AssistedL1T3Vendor self-reportGoogle Research shows Google Earth AI can help forecast disease outbreaks source2026-10-06
L2Partial automationL2T3Vendor self-reportGoogle DeepMind launches EmbeddingGemma 2 multimodal on-device embedding model source2026-10-06
L2Partial automationL2T3Vendor self-reportGoogle Earth AI Population Dynamics Foundation Model case studies source2026-10-06
L1AssistedL1T3Vendor self-reportGoogle Research introduces TEE-based federated learning system source2026-10-02
L2Partial automationL2T3Vendor self-reportGoogle DeepMind launches SynthID Bio watermarking for biological designs source2026-10-01
L2Partial automationL3↓ T3 allows only L2T3Vendor self-reportGoogle science AI model ranks highest in CDC flu forecasting source2026-09-30
L2Partial automationL2T3Vendor self-reportMicrosoft Research machine learning system predicts space-weather damage source2026-09-30
L1AssistedL1T3Vendor self-reportGoogle Research introduces Diffusion Controller for image generation steering source2026-09-29
L1AssistedL1T3Vendor self-reportMicrosoft Research introduces Quine, a multimodal world model of biology source2026-09-29
L2Partial automationL3↓ T3 allows only L2T3Vendor self-reportQoder knowledge engine builds knowledge graph from code source2026-09-28
L1AssistedL1T3Vendor self-reportAnthropic Science Blog: Claude can do Nine Loops source2026-09-25
L2Partial automationL2T3Vendor self-reportGoogle announces unified multi-agent framework for long-form video source2026-09-24
no update mentioned itL0T3Vendor self-reportTencent Hunyuan research extends critical-batch-size theory to online LLM RL source2026-09-24
L1AssistedL1T3Vendor self-reportMicrosoft Research findings on moving robot AI inference off-device source2026-09-23
L2Partial automationL2T3Vendor self-reportTencent Hunyuan introduces WebCraftBench for agent web-app evaluation source2026-09-22
L1AssistedL1T3Vendor self-reportMicrosoft Research highlights RetroChimera model in Nature paper source2026-09-21
L2Partial automationL3↓ T3 allows only L2T3Vendor self-reportAlibaba releases Qwen3.8-LiveTranslate real-time interpretation model source2026-09-19
L1AssistedL1T3Vendor self-reportGoogle Research introduces MilleMiglia logistics benchmark source2026-09-18
L1AssistedL1T3Vendor self-reportAnthropic work on running biology open-source models more cheaply source2026-09-17
L1AssistedL1T3Vendor self-reportGoogle Research introduces Retrieve-for-Train framework source2026-09-15
L1AssistedL1T3Vendor self-reportGoogle publishes white paper on AI for crisis resilience source2026-09-15
L2Partial automationL2T3Vendor self-reportTencent Hunyuan introduces EvolveScaler source2026-09-15
L2Partial automationL3↓ T3 allows only L2T3Vendor self-reportMiniMax H3 achieves over 2x real-time denoising on 8x B200 source2026-09-14
L2Partial automationL2T3Vendor self-reportGoogle Antigravity launches AlphaGenome Atlas Skills and /boost command source2026-09-11
L2Partial automationL2T3Vendor self-reportGoogle DeepMind uses AI to reconstruct memory for documentary 'Love, Rendered' source2026-09-11
L1AssistedL1T3Vendor self-reportGoogle Research TIPSv2 demo and Q&A at ECCV 2026 source2026-09-11
L2Partial automationL2T3Vendor self-reportGoogle Research introduces ToolGrad framework for tool-use datasets source2026-09-10
L1AssistedL1T3Vendor self-reportGoogle DeepMind researcher receives PAMI Mark Everingham Prize at ECCV 2026 source2026-09-10
L2Partial automationL2T3Vendor self-reportTencent Hunyuan releases AuK open-source speech model source2026-09-10
L1AssistedL1T3Vendor self-reportOpenAI shares cyber defense architecture and playbook source2026-09-09
L2Partial automationL2T3Vendor self-reportGoogle Research releases complete male fruit fly brain wiring diagram source2026-09-09
L2Partial automationL3↓ T3 allows only L2T3Vendor self-reportGoogle DeepMind launches AlphaGenome Atlas source2026-09-08
L2Partial automationL3↓ T3 allows only L2T3Vendor self-reportMeta's AIRA₃ competes in Kaggle challenge to fine-tune Nemotron 30B source2026-09-05
L2Partial automationL3↓ T3 allows only L2T3Vendor self-reportClaude completes first formalized proof of Fermat's Last Theorem source2026-09-04
L1AssistedL1T3Vendor self-reportGoogle Research evaluates cross-population genetic risk prediction methods source2026-09-03
L2Partial automationL2T3Vendor self-reportGoogle DeepMind introduces WeatherNext 3 source2026-09-03
L2Partial automationL2T3Vendor self-reportGoogle discusses WeatherNext forecasting models source2026-09-02
L2Partial automationL2T3Vendor self-reportMiniMax reports emergent world-model behavior in H3 source2026-09-02
L2Partial automationL3↓ T3 allows only L2T3Vendor self-reportDreamX-Creator audio-visual generation model source2026-09-02
L2Partial automationL2T3Vendor self-reportGoogle Research presents MAPL-EMIT methane tracking model source2026-09-01
L2Partial automationL3↓ T3 allows only L2T3Vendor self-reportHy4 preview autonomously finds inference bottlenecks, lifts throughput 31.8% source2026-09-01
L1AssistedL1T3Vendor self-reportAnthropic research: Training a Misaligned Reward Seeker source2026-09-01
L2Partial automationL2T3Vendor self-reportGoogle Research introduces TimesFM-3 time series foundation model source2026-08-31
L1AssistedL1T3Vendor self-reportGigaPath-Flash and GigaTIME-Flash pathology foundation models source2026-08-31
L1AssistedL1T3Vendor self-reportTencent Hunyuan compresses Hy4-preview to ~200GiB GGUF with MIX-STQ1_0 source2026-08-29
L2Partial automationL3↓ T3 allows only L2T3Vendor self-reportAnthropic Fellows research: Claude autonomously aligns other AIs source2026-08-28
L1AssistedL1T3Vendor self-reportAntigravity Teamwork used for research breakthroughs source2026-08-27
L2Partial automationL2T3Vendor self-reportGoogle Research introduces planetary prediction engine source2026-08-27
L2Partial automationL2T3Vendor self-reportGoogle Research introduces GlucoFM glucose monitoring foundation model source2026-08-26
L2Partial automationL2T3Vendor self-reportGoogle Research presents AgentHands XR prototype source2026-08-25
L1AssistedL1T2Independently corroboratedNVIDIA SANA Sol Engine cuts MiniMax H3 768p latency to 14.93s source2026-08-24
no update mentioned itL0T3Vendor self-reportMicrosoft Research releases Skala 1.1 deep-learning functional source2026-08-24
L2Partial automationL3↓ T3 allows only L2T3Vendor self-reportGitHub adds context-aware secret classifier checking candidates under 2ms sourceno update mentioned it
Build research prototypes and run comparative experiments43 evidence12 tasksL2Partial automation

Open this work →

PublishedClaimedEvidence tierUpdateWhen
L2Partial automationL3↓ T3 allows only L2T3Vendor self-reportAMIE matches doctor diagnoses in 90% of 100 patient interactions with 0 safety stops source2026-10-08
L2Partial automationL2T3Vendor self-reportGoogle's AMIE studied prospectively in real clinic, results in The Lancet source2026-10-08
L1AssistedL1T3Vendor self-reportAstrophysicist uses Claude Science to build first complete ultraviolet sky map source2026-10-08
L1AssistedL1T3Vendor self-reportGitHub releases ReviewBench for AI code review agents source2026-10-06
L1AssistedL1T3Vendor self-reportGoogle Earth AI Population Dynamics Foundation Model case studies source2026-10-06
L2Partial automationL2T3Vendor self-reportGoogle science AI model ranks highest in CDC flu forecasting source2026-09-30
L2Partial automationL2T3Vendor self-reportMicrosoft Research machine learning system predicts space-weather damage source2026-09-30
no update mentioned itL0T3Vendor self-reportTencent Hunyuan releases ExplorationBench benchmark source2026-09-30
L1AssistedL1T3Vendor self-reportGoogle Research introduces Diffusion Controller for image generation steering source2026-09-29
L1AssistedL1T3Vendor self-reportMicrosoft Research introduces Quine, a multimodal world model of biology source2026-09-29
L1AssistedL1T3Vendor self-reportAnthropic Science Blog: Claude can do Nine Loops source2026-09-25
L1AssistedL1T3Vendor self-reportProject Suncatcher to test TPUs in orbit source2026-09-24
L2Partial automationL2T3Vendor self-reportGoogle announces unified multi-agent framework for long-form video source2026-09-24
L1AssistedL1T3Vendor self-reportTencent Hunyuan research extends critical-batch-size theory to online LLM RL source2026-09-24
L1AssistedL1T3Vendor self-reportOpenAI releases MentalHealthBench source2026-09-23
L2Partial automationL2T3Vendor self-reportTencent Hunyuan introduces WebCraftBench for agent web-app evaluation source2026-09-22
L1AssistedL1T3Vendor self-reportMicrosoft Research highlights RetroChimera model in Nature paper source2026-09-21
L1AssistedL1T3Vendor self-reportGoogle Research introduces MilleMiglia logistics benchmark source2026-09-18
L2Partial automationL2T3Vendor self-reportMuse Voice Transcribe achieves lowest semantic WER in Pipecat benchmark source2026-09-14
L2Partial automationL2T3Vendor self-reportGoogle Antigravity launches AlphaGenome Atlas Skills and /boost command source2026-09-11
L2Partial automationL2T3Vendor self-reportGoogle DeepMind uses AI to reconstruct memory for documentary 'Love, Rendered' source2026-09-11
L2Partial automationL2T3Vendor self-reportGoogle Research introduces ToolGrad framework for tool-use datasets source2026-09-10
L2Partial automationL2T3Vendor self-reportMeta's AIRA₃ competes in Kaggle challenge to fine-tune Nemotron 30B source2026-09-05
L1AssistedL1T3Vendor self-reportMicrosoft reports HydraFusion results on Terminal-Bench 2.1 source2026-09-04
L1AssistedL1T3Vendor self-reportGoogle Research evaluates cross-population genetic risk prediction methods source2026-09-03
L2Partial automationL2T3Vendor self-reportGoogle DeepMind introduces WeatherNext 3 source2026-09-03
L2Partial automationL2T3Vendor self-reportQwen introduces E-Commerce Bench benchmark source2026-09-03
L2Partial automationL2T3Vendor self-reportMiniMax reports emergent world-model behavior in H3 source2026-09-02
L1AssistedL1T3Vendor self-reportQwen3.8-Max-0902 released source2026-09-02
L2Partial automationL2T3Vendor self-reportGoogle Research presents MAPL-EMIT methane tracking model source2026-09-01
L1AssistedL1T2Independently corroboratedLatchBio evaluates Grok 4.6 on biosecurity tasks source2026-09-01
L1AssistedL1T3Vendor self-reportAnthropic research: Training a Misaligned Reward Seeker source2026-09-01
L2Partial automationL2T3Vendor self-reportGoogle Research introduces TimesFM-3 time series foundation model source2026-08-31
L1AssistedL1T3Vendor self-reportGigaPath-Flash and GigaTIME-Flash pathology foundation models source2026-08-31
L1AssistedL1T3Vendor self-reportOpenAI launches Rosalind Workbench source2026-08-28
L2Partial automationL3↓ T3 allows only L2T3Vendor self-reportAnthropic Fellows research: Claude autonomously aligns other AIs source2026-08-28
L1AssistedL1T3Vendor self-reportAntigravity Teamwork used for research breakthroughs source2026-08-27
L2Partial automationL2T3Vendor self-reportGoogle Research introduces planetary prediction engine source2026-08-27
L1AssistedL1T3Vendor self-reportGoogle Research introduces GlucoFM glucose monitoring foundation model source2026-08-26
L1AssistedL1T3Vendor self-reportGoogle Research presents AgentHands XR prototype source2026-08-25
L2Partial automationL2T3Vendor self-reportWan 3.0 ranks highly in Koyal Film Arena animated video categories source2026-08-25
L1AssistedL1T2Independently corroboratedNVIDIA SANA Sol Engine cuts MiniMax H3 768p latency to 14.93s source2026-08-24
L2Partial automationL2T3Vendor self-reportGemini ML skills write and test PySpark code in notebooks sourceno update mentioned it

Occupations this market is served by

One market is served by several occupations, so these are not shares of a job.

OccupationConfidence
Computer and Information Research Scientists0.97
Bioinformatics Scientists0.85
Computer Science Teachers, Postsecondary0.84
Engineering Teachers, Postsecondary0.81
Robotics Engineers0.81
Remote Sensing Scientists and Technologists0.79
Data Scientists0.78
Education Teachers, Postsecondary0.72
Operations Research Analysts0.65
Postsecondary Teachers, All Other0.62
Blockchain Engineers0.62

Where this page's numbers come from

Snapshot
snapshot-1
Generated
2026-10-09 08:10 UTC
Ontology
1.0.0
Ontology schema
2.4.0
Method version
activity-level-1
Content SHA-256
0b46df0246af8a1dc3fd9ce8e6ac9812925ace553984d2b7d45d363d32e27681
Rows in this snapshot
10,482 work-to-task edges · 900 updates · 333 models · 58 topics · 1,888 market-to-occupation edges · 375 source accounts · 1 coverage record · 1 progress record · 11 non-technical conditions · 401 updates that landed on work · 1,102 evidence · 614 work in markets · 506 non-technical condition-to-work edges