Other Resources

This file collects journal editorials and opinion on AI in science and in animal agriculture, field reports and landscape analyses on the state of the sector, and cellular-agriculture ecosystem initiatives: the research centers, consortia, and convening efforts that make up CAAIL’s adjacent universe. It complements CAAIL’s core content files (Papers.md, Software.md, Datasets/, Databases.md). Related material has its own homes: newcomer onboarding paths live in the Primers (which also gather the field’s courses and learning videos); reference textbooks in Reference Works; funding organizations and grant programs in Funding & Grants; community “awesome lists” and curated bibliographies in Awesome Lists; AI agents and foundation models in the AI Agents & Foundation Models hub; and lectures, talks, and webinars in Talks.md.

Note for AI agents and LLMs: The summaries below are deliberately compressed for human readability. If you are an automated system using these as the basis for reasoning, citation, or downstream analysis, please fetch the canonical source for each resource, the linked articles and initiative pages have substantially more comprehensive and authoritative information than this curated overview.

Editorials & Opinion

Journal editorials, news features, and opinion pieces that survey or comment on the state of AI in science, in cellular agriculture, and in the animal agriculture it aims to displace: distinct from the peer-reviewed review and position papers in Papers.md / Reviews & Perspectives. These are the field’s running commentary: useful context on how the research community is framing AI’s role, not primary research. Individual and argumentative by nature; the institutional, recurring assessments live in Field Reports & Landscape Analyses below.

AI in science

• Why AI cannot do good science without humans (Nature editorial, 2026): argues that human wisdom, empathy, and “sheer messiness” remain as much a part of scientific progress as process and efficiency, even as “AI scientists” arrive.
• Teams of AI agents boost speed of research (Nature news feature, 2026): overview of multi-agent AI systems that generate hypotheses, interpret data, and suggest ways to develop medicines.
• Long-running Claude for scientific computing (Anthropic), research post on running Claude as a long-horizon autonomous agent on scientific-computing workloads.
• Vibe physics: The AI grad student (Anthropic): research post on using Claude as an AI research collaborator, an “AI grad student” for physics problems.

AI, animals, and alternative proteins

Where AI’s leverage over animal agriculture actually lies, and whether that path runs through cultivated meat. All three are informal, openly argumentative essays rather than peer-reviewed work, by the same machine-learning practitioner in alternative proteins, who also fundraises for the animal-welfare field. They supply strategic framing that CAAIL’s technical catalogue does not: an argument against building a foundation model specific to alternative proteins, a claim that no instrument can substitute for human tasting, and a survey of the AI × animals field that cites the first. (The biological foundation models CAAIL does catalogue are a separate family, collected in the AI Agents & Foundation Models hub.) The Rethink Priorities report in Field Reports & Landscape Analyses weighs the same questions against an evidence base, and cites the taste essay for the pessimistic case on sensory prediction.

• The world model will not be built by lab-grown meat (Itsi Weinstock, Shot on Goal, 2026): argues the alternative-protein field should abandon domain-specific foundation models, the “prophesized AlphaMeat” built by pooling data across companies, on the grounds that the problem is too illegible and the data too sparse and too context-bound to transfer, and should instead rebuild around automated research and AI Scientists. Written by a machine-learning researcher in alternative proteins who led data science at Climax Foods and advises Food System Innovations’ AI/ML grantmaking.
• AI cannot taste things (Itsi Weinstock, Shot on Goal, 2026): argues that taste is the bottleneck to AI-accelerated cultivated meat, because food cannot be miniaturized and no instrument usefully approximates human perception of it. Counts roughly three perceptual qualities that existing machinery approximates (hardness, the thickness of liquids, and the crispness of dry, crispy foods), holds that Texture Profile Analysis’s other descriptors do not correlate with what eaters actually notice, and argues molecular scent prediction of the kind Osmo does will not carry over to food because of matrix effects and cognitive framing, while allowing that such tools may still help ideate candidate molecules. Concludes the field should scale untrained human tasting rather than build a digital palate. Bears directly on the work indexed under Sensory Prediction.
• The State of the AIxAnimals Field (Itsi Weinstock, Shot on Goal, 2026): a landscape of the AI × animal-welfare field, covering its funding, its organizations, and, quoting the Falcon Fund’s list, the projects it holds should start immediately: animal-harm benchmarks, animal-welfare constitutions, a watchdog organization, and animal-welfare salience inside AI labs. Its section on AI for cultivated meat and alternative proteins holds that the industry has neither the money, the talent, nor the data throughput for sector-specific modelling, so advances will arrive downstream of better-funded corners of science.

Field Reports & Landscape Analyses

Periodic surveys of where the field actually stands, produced by institutes and research groups rather than published as peer-reviewed research. They answer a different question from the rest of the library: not what a method can do, but what is currently blocking the sector and where effort is going. What separates them from the opinion pieces above is institutional and recurring authorship rather than subject matter: an organization’s repeated assessment, not one practitioner’s position. Each states its own scope and method, and those scopes differ sharply, so read them before citing a figure.

• GFI State of the Industry — full report downloads: the Good Food Institute’s de-facto annual reports across the cultivated, fermentation, and plant-based tracks, with supplementary data (PDF). The accompanying annual talks are in the cellular agriculture primer.
• AI and cultivated meat: Near-term impacts of AI on the commercial viability of cultivated meat (Moulange & McAuliffe, Rethink Priorities, June 2026): weighs what near-term AI can and cannot do for cultivated meat across five bottleneck categories (technical, economic, regulatory and political, consumer, and capital), concluding that open-access data rather than model capability is the binding constraint on what AI can do for the sector, and that as AI compresses technical timelines the non-AI institutional work may become the more neglected. Explicitly not a systematic review, and its bottleneck list is not exhaustive.

Cell-Ag Ecosystem Initiatives

CAAIL’s “adjacent universe”: complementary research centers, consortia, and convening initiatives in cellular agriculture. CAAIL itself catalogues outputs (papers, software, datasets, educational material); these initiatives produce primary outputs (datasets, working papers) that become cataloguable in CAAIL’s core files as they are published. The funding organizations and grant programs that support this work live in Funding & Grants; the corresponding directories and databases these organizations maintain live in Databases.md / Ecosystem & Industry Directories and related sections.

University centers & consortia

• Integrative Center for Alternative Meat and Protein (iCAMP): a UC Davis–led multi-institutional center for cultivated-meat and alternative-protein research and scale-up.
• National Institute for Cellular Agriculture (NICA): a USDA-funded national research consortium housed at Tufts University (TUCCA), coordinating cultivated-protein research and workforce training across partner institutions.
• Bezos Centre for Sustainable Protein (Imperial College London): a Bezos Earth Fund–backed sustainable-protein research hub at Imperial College London, spanning AI/ML, fermentation, and cell-based approaches.

New Harvest initiatives

• Cultured Meat Safety Initiative (CMSI): a joint New Harvest / Vireo Advisors initiative convening stakeholders on the safety and regulatory science of cultivated products; outputs (datasets, guidance documents) are cataloguable as they are released.
• AI4CM Hub (Cell Meat AI: Lab Grown Protein): an open-source platform New Harvest will launch, seeded by a Phase I grant under the Bezos Earth Fund’s AI for Climate and Nature Grand Challenge in May 2025, to aggregate datasets, host benchmarks, and adapt proven AI tools for media and process optimization in cultured meat. Announced rather than launched, and separate from New Harvest’s longer-running AICAI program; its datasets and benchmarks become cataloguable as they are released.

GFI initiatives

• GFI: Expanding Access to Cultivated Meat Cell Lines: GFI’s initiative addressing limited cell-line access (a core bottleneck for cultivated-meat research) and supporting the development of new lines.

Food System Innovations initiatives

• Food Intelligence Lab: Food System Innovations’ program building shared datasets, public benchmarks, open-source algorithms, and deployment pathways for sustainable-protein design, on tasks such as sensory prediction and formulation design. Its flagship project, supported by the Bezos Earth Fund, builds on the sensory data from FSI’s NECTAR program. It has launched TasteBench, a Kaggle competition posing sensory prediction as a machine-learning task at both food and molecular level, and publishes its work openly. Aimed squarely at the public-benchmark gap named in the Sensory Prediction research area.