Blog
Technical deep-dives on AI peer review, research gap discovery, and open-access publishing. Written by the team behind Science AI Journal.
- research-gaps
The open problem in Mediterranean agroforestry: who is running the soil?
Thirty-nine recent papers agree that soil microbial communities are central to Mediterranean agroforestry performance — but none has resolved what drives their composition. Here is what the evidence shows, and what the field still needs.
- ai-peer-review
How AI Review's Agents Are Calibrated on 69,000 Real Peer Reviews
AI Review's seven language-model agents read real peer-review examples from a 69,000+ record corpus collected from 19 open-review platforms. Here is how the examples are chosen, and what they cannot do.
- ai-peer-review
When not to use AI peer review: honest limits and edge cases
Our 8 AI agents are calibrated on 69,000 real peer reviews, but some papers fall outside their scope. Here is what to watch for before running a review.
- research-gaps
The open problem in gravitational physics: developing and testing the Gravitational Saturation Theory
Across 29 papers published in 2026, researchers converge on the same gap: the Gravitational Saturation Theory needs rigorous experimental verification and formal extension. Here is what the literature says, what remains unresolved, and what it would take to move forward.
- research-gaps
The open problem in chemistry: why a shared scientific vocabulary remains out of reach
Twenty-five papers published in or around 2026 independently reach the same diagnosis: chemistry's knowledge base is fracturing along subfield, methodological, and national lines, with no integrative institutions, no shared vocabulary, and no transparent synthesis practices to hold it together.
- ai-peer-review
Hallucination Guardrails for an AI Reviewer: What We Do, What We Still Can't Fix
How Science AI Journal's 8 AI agents guard against fabricated review claims -- and where those guards have real limits. A transparent technical account.
- research-gaps
The open problem in teacher education: why professional development keeps failing at the policy level
Thirty-three studies spanning Southeast Asia and beyond agree: teachers lack sustained, well-resourced professional development — but no integrated policy framework exists to fix it.
- research-gaps
The open problem in biochar agriculture: how much is enough, and what should it be mixed with?
Forty-three recent studies confirm that biochar improves soil health and crop yields — but none has systematically resolved the optimal biochar-to-vermicompost ratio for high-yield growing substrates. Here is what the evidence shows, and what the field still needs.
- ai-peer-review
How AI enforces CONSORT, STROBE, and PRISMA in peer review
How our methodology agent checks RCTs against CONSORT 2010, observational studies against STROBE, and systematic reviews against PRISMA, calibrated on 69,000 real peer reviews.
- research-gaps
The open problem in gravitational wave physics: what 23 recent papers still cannot answer
Since 2015, gravitational wave detectors have transformed astrophysics. But 23 papers published in 2025–2026 converge on a cluster of deep, unresolved questions about black holes, spacetime, and dark matter that current observations leave open.
- research-gaps
The open problem in AI education: technology alone is not enough
Thirteen recent studies agree: deploying AI tools in classrooms produces no reliable learning gains unless the design is pedagogically intentional. Here is what the research says — and what still needs to be resolved.
- ai-peer-review
The complete guide to AI peer review in 2026
How 8 specialized AI agents, calibrated on 69,000 real peer reviews from 19+ platforms, deliver rigorous, discipline-specific feedback in under 15 minutes.
- research-gaps
The open problem in inclusive education: Why teacher training still falls short of the classroom
18 recent studies converge on the same finding: teachers enter inclusive classrooms without the practical preparation they need. Here is what the literature shows, what remains unresolved, and what it would take to close this gap.
- perspective
Every Tool That Made Scholarship Faster Was Called Cheating First
From Trithemius attacking the printing press in 1492 to today's anxieties about AI, every technology that compressed the mechanical labour of research was called cheating first — then became invisible infrastructure. Where AI honestly belongs in that lineage, and where the line has to be drawn.
- research-gaps
The open problem in AI-driven education: What 33 studies still couldn't answer
A cluster of 33 recent papers on AI in education converges on the same three unresolved questions: long-term impact, cross-cultural validity, and algorithmic fairness. Here is what the literature says — and what it leaves open.
- guides
How Long Does Peer Review Take? (And How to Get Editor-Ready Feedback in 15 Minutes)
Peer review typically takes 1 to 6 months to the first decision, and often longer. Here is what drives the timeline, realistic ranges by field, and how to catch the problems reviewers will flag before you submit.
- guides
How to Check if a Journal Is Predatory: A 5-Minute Checklist
A predatory journal charges publication fees while skipping real peer review. Here is a fast, evidence-based checklist to vet any journal before you submit — plus a tool that flags them automatically.
- guides
Research Gap Examples by Field (With How to Find Your Own)
Concrete research gap examples from medicine, computer science, education, psychology, engineering, and biology — and a repeatable method for finding one in your own field.
- guides
How to Find a Research Gap: A Practical Guide (With Examples)
A research gap is an unanswered question the existing literature has not resolved. Here are the seven types, how to find one by hand, and how the guided finder does it, table first.
- research-gaps
Measuring AI's Impact on Student Learning: Open Questions in Education Assessment
While AI integration in higher education is expanding rapidly, critical gaps remain in measuring pedagogical impacts—from domain-specific cognitive outcomes to academic integrity safeguards.
- research-gaps
Bridging Innovation and Clinical Evidence in Modern Healthcare
Exploring critical research gaps in translating emerging medical technologies, novel treatments, and AI-assisted diagnostics into clinical practice—from validation studies to real-world implementation challenges.
- research-gaps
Who Was Not in the Study? Open Questions in Clinical Research Generalizability and Causal Inference
Most clinical findings are published before the question of generalizability has been answered. We trace six specific gaps where promising relationships between biomarkers, infections, and functional outcomes cannot yet be trusted beyond their original study cohort.
- research-gaps
Measuring What AI Actually Does to Learning: Six Open Questions
As AI tools flood classrooms from primary school to postgraduate research, the field lacks standardized protocols to detect whether algorithmic assistance builds or displaces genuine understanding. We map six research gaps drawn from 2024–2026 primary literature.
- research-gaps
Thermal and Environmental Durability in Emerging Materials: Seven Open Questions
From ultrathin 2D waveplates to stretchable OLEDs and high-entropy nanoalloys, the 2020s materials renaissance shares a common blind spot: nobody has tested what happens next year. We map seven open durability questions drawn from recent primary literature.
- engineering
Peer review in 15 minutes: how Science AI Journal works
An inside look at our 8-agent review engine: what each agent checks, why eight narrow reviewers beat one broad prompt, and what the report does not claim.
- engineering
Detecting prior publication across 8 sources in under 12 seconds
How we fan out across CrossRef, PubMed, Europe PMC, Unpaywall, arXiv, medRxiv, bioRxiv, and a local 4.5M-paper FTS5 index to catch prior publication early.
- research
How we mined 120,000+ research gaps from the literature
How our gap finder mines the limitations authors write about their own work, and why most 'AI gap-finders' hallucinate.