New Frontiers in AI Research: Infant Cognition, Brain Analysis, Drug Discovery, and Quantum Computing

연구/벤치마크 | Thu Jul 16 2026 00:00:00 GMT+0000 (Coordinated Universal Time) | 4 sources

Frontier AI research is pushing boundaries in human intelligence and life sciences.

Analysis

[Meta, Stanford, University of Tokyo, École Normale Supérieure] released the EgoBabyVLM benchmark ^[1]

• Evaluates vision language models based on 1
• 000 hours of infant headcam footage
• State-of-the-art VLMs failed to process baby-perspective data
• Highlights the importance of multimodal and tactile learning

[BabyLM Challenge] evaluated language learning ability with limited data ^[1]

• Uses training data equivalent to a 10-year-old child's exposure (tens of millions of words)
• Demonstrated that transformer-based models can learn grammar
• Challenges Chomsky's theory of language acquisition

[Hemispheric] developed a frontier AI model for brain activity analysis ^[4]

• Raised $52M in funding
• Collected 250
• 000 hours of brain data from 100
• 000 people
• Diagnoses via 15-minute EEG headset measurements
• Targets FDA approval for PTSD product in 2027

[Miles Wang] moved to launch an AI drug discovery startup ^[3]

• In talks for $200M investment at $2B valuation
• Lightspeed in discussions to lead the round
• Targets discovering new uses for existing FDA-approved drugs
• OpenAI researchers expected to join

[PsiQuantum] leads the race to build practical quantum computers ^[2]

• Founded in 2016 by four British physicists
• Photon-based quantum computing architecture
• Envisions thousands of chips housed in 100 stainless-steel cabinets

Sources

• [1] AI Isn’t Smarter Than a Baby—Yet - Wired AI
• [2] The Download: a useful quantum machine and a record-breaking subsea tunnel - MIT Technology Review AI
• [3] OpenAI researcher Miles Wang in talks to launch AI drug discovery startup valued at $2B - TechCrunch AI
• [4] An Inventor of Apple’s FaceID Wants to Analyze Your Brain’s Health With AI - Wired AI