🔍 Read the full analysis: A Guide To Google And CSIRO’s AI Research On Tasmania’s Giant Kelp on ThorstenMeyerAI.com
Get hardware and tech essentials delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Google says CSIRO used a version of its DeepVariant AI, trained for giant kelp DNA, to analyze 600 wild specimens from 40 Tasmanian coastal sites. The work is intended to help scientists select kelp for selective breeding; the analysis and reported genetic isolation have not yet been supported by a published paper or dataset, and survival in warmer seas remains untested.
Google said on November 12 that Australia’s national science agency, CSIRO, used a version of Google’s AI variant-calling tool, DeepVariant, trained for kelp DNA to analyze 600 wild giant kelp specimens collected from 40 sites along Tasmania’s coast. The analysis is intended to help scientists select plants for selective breeding as they work to restore kelp forests that, according to CSIRO, have declined by about 95%; whether bred kelp can withstand warmer ocean conditions has not yet been tested in the sea.
Google’s announcement says the researchers adapted DeepVariant, whose standard models are trained on human data, to analyze genetic differences in the kelp specimens. Google said the work found that remaining Tasmanian kelp patches are genetically isolated. That result, along with the 600-specimen analysis, is so far described in Google’s post and video; the source material reports that no corresponding paper or dataset has been made available yet.
A separate part of the work is documented in a peer-reviewed paper published in BMC Genomics in June 2026, authored by CSIRO scientists and a Google researcher. It presents two reference genomes assembled from Australian giant kelp, one from Victoria and one from Tasmania. The paper reports that the genomes differ from each other by 0.2%, while each differs from the Californian reference by about 1.5%—a reason researchers say a local reference can improve analysis of Australian samples.
Google’s accompanying video says its open-source tools DeepConsensus and DeepPolisher were used to correct errors in raw genetic reads during genome assembly. The next proposed step is to cross genetically distinct native survivors in a laboratory and grow young kelp for planting on reefs. The source describes this as selective breeding, not genetic modification; it does not report that restoration plantings have already demonstrated heat tolerance.
Giant kelp · Tasmania · CSIRO with Google AI
From the reef to the lab and back: where the AI fits
Five steps, as described by CSIRO and Google. The AI works in steps 2 and 3; people do the diving, the lab work and the planting.
1Reef
Collect wild kelp
600 wild specimens from 40 sites along the Tasmanian coast, according to Google.
2Lab + AI
Build a local reference
Long-read DNA plus Google’s DeepConsensus and DeepPolisher: two Australian reference genomes, published in BMC Genomics.
AI step3AI
Read each plant
A version of DeepVariant trained for kelp finds where each specimen’s DNA differs.
AI step4Lab
Choose and cross
Distinct survivors with genetic variation and heat-tolerant traits are crossed. Selective breeding, not genetic modification.
5Reef
Replant and watch
Young kelp is attached to rocky reef in pilot tests. Warm summers show which plants survive.
Sources: Google blog post, Nov 12, 2026, and Google Research video; Scharfenstein et al., BMC Genomics, 2026; CSIRO, 2024, and CSIRO on Google Arts & Culture. The scene is a schematic, not a map or real data.
Graphic: Powered by Thorsten Meyer AI
Using Genomics to Guide Kelp Breeding
The project addresses a severe ecological decline with a specific research method: identify genetic variation in surviving kelp, then use that information to guide breeding. Giant kelp forests provide habitat and nursery areas for marine life, and CSIRO says the wider ecosystem depends on them. The goal is not for AI to restore the forest directly, but to support decisions about which native plants scientists might cross.
Building Australian reference genomes also matters because the earlier comparison genome came from California. The reported genetic difference between that specimen and the Australian genomes suggests that using the Californian sequence alone could make it harder to interpret variation in local kelp. Better-matched reference data may help researchers study the remaining populations, though it does not by itself show that a particular kelp will survive heat or other pressures.
The distinction between laboratory analysis and ecological results is central. The announcement describes a tool for reading DNA and identifying differences; it does not establish that AI has made kelp heat-resistant, that breeding will succeed, or that restored forests will recover. The practical test comes when bred kelp is grown and planted in the ocean.
Products Worth Considering
Logitech MX Keys S Combo Wireless Keyboard and Mouse, Graphite
Apple MacBook Pro 16-inch (2021), M1 Pro, 10-core CPU, 16-core GPU, 16GB RAM, 512GB SSD, Silver
LG UltraGear evo 52-Inch 5K2K 240Hz Curved Gaming Monitor
As an affiliate, we earn on qualifying purchases.
Why Tasmania’s Kelp Is Declining
Giant kelp, or Macrocystis pyrifera, is a large brown seaweed that forms underwater forests in cold, shallow waters. CSIRO says it can grow to around 30 metres in Tasmanian waters. Those forests support a range of marine species, and the agency describes giant kelp forest as an endangered marine community type.
CSIRO attributes the decline around Tasmania to warming water, nutrient-poor currents and sea urchins. It says the East Australian Current has pushed farther south, bringing warmer, less nutrient-rich water and urchins into the region; warming in the Tasman Sea is reported to be about twice the global average. The source material says CSIRO estimates a 95% decline in the kelp populations, while a cited preprint, drawing on earlier research, gives an estimate of 95–98% loss of East Tasmania kelp cover since 1940.
Google and CSIRO’s broader collaboration was announced in February 2024 with the University of Tasmania’s Institute for Marine and Antarctic Studies, The Nature Conservancy, the Great Southern Reef Foundation and the Kelp Forest Alliance. Its work included a mapping strand using satellite imagery and Google tools. The November 2026 announcement concerns the genomics strand, not a report that mapping or restoration has returned kelp to depleted reefs.
“trained a version of our AI model DeepVariant to analyze our wild samples”
— Google, in a video accompanying its November 12 announcement
Evidence Still Awaiting Publication
The 600-specimen analysis and Google’s claim that remaining kelp patches are genetically isolated have not, in the supplied source material, been backed by a peer-reviewed paper or a public dataset. The details needed to assess the analysis—including its methods, the precise measure of isolation and how results vary among sites—are not provided in the announcement as summarized here. The published genome paper is a separate, peer-reviewed result and should not be treated as independent verification of every claim about the later sample analysis.
It is also not yet clear how many crosses will be made, which genetic traits will guide breeding, when young kelp will be planted, or how success will be measured. Most importantly, no reported field result establishes that bred kelp survives warmer summers or can rebuild forests under the conditions linked to the decline.
From Lab Crosses to Reef Trials
According to the project description, researchers plan to use the genetic information to cross distinct native survivors in the laboratory, then grow young kelp for planting on reefs. The source does not give a firm schedule for those steps or specify a planting site. Further publications or released datasets could allow independent review of the 600-specimen findings.
The key milestone will be evidence from the ocean: whether planted kelp establishes, persists through warm seasons and contributes to forest recovery. Until such results are reported, the AI work is best understood as research intended to inform breeding—not proof that restoration has succeeded.
Key Questions
What did Google and CSIRO announce?
Google said CSIRO used a version of DeepVariant adapted for kelp DNA to analyze 600 wild specimens from 40 Tasmanian coastal sites. The work is intended to inform selective breeding and restoration research.
Has the AI restored Tasmania’s kelp forests?
No restoration result is reported. The AI analyzes genetic data; scientists plan to use their findings to help select kelp for laboratory crosses and eventual reef planting.
Is the 600-specimen analysis peer reviewed?
The source material says the analysis and reported genetic isolation are described in Google’s announcement and video, with no related paper or dataset identified there. A separate paper on two Australian reference genomes was published in BMC Genomics in June 2026.
Is the planned breeding genetic modification?
The project describes crossing distinct native kelp in the lab as selective breeding, not genetic modification.
What is the main test still ahead?
Researchers still need to establish whether bred kelp can grow and survive in the sea, including through warmer summers. The supplied material reports no field result answering that question.
Source: ThorstenMeyerAI.com
