By Tanveer Singh
Imagine waking up one morning to hear the cheerful sounds of bird songs, only to realize that the bird you loved is no longer there – it has vanished forever. The same could be true for the animal you are deeply connected to, that symbolizes your identity or spirit, lost to extinction and removed from the natural world. This heartbreaking decline has become more common in recent decades, as countless birds, from the passenger pigeon to the Spix’s macaw, and animals, such as the Asiatic cheetah and the northern white rhino, have disappeared from the world due to habitat loss, climate change, and human activity. Even the sparrows that once filled our childhood with wonderful memories are now a rare sight because of the construction of buildings, towers, and the use of pesticides.
In response to these losses, scientists are constantly working on the new field: De-extinction or species revival that uses genetic engineering and cloning, not only to make the DNA of lost species but also to simulate their behaviors and roles within ecosystems. However, the question of deep curiosity is what if recent Artificial Intelligence technologies are applied to de-extinction? Will AI be able to make what was once impossible, and how efficiently will it be able to preserve the extinct animals?
To respond to your curiosity about how the process of the revival of species works, here is the answer. The procedure starts with the deep research of fossil fuels or fragments of ancient DNA from museum specimens or archaeological finds. They served as a blueprint, and then traditional methods like cloning and selective breeding were used to make these species. However, these methods are often hindered because of contamination or missing DNA fragments or loss of some pieces of fossil fuels, which makes it difficult to recreate a full genome (a complete set of genetic information). Take the example of the Moa Bird of New Zealand; the bird was known to be extinct in the mid-15th century, and when fossil remains of it were gathered, it was found that there were many missing and damaged DNA parts; as a result, the full genome could not be constructed through cloning. These challenges underscore the need for artificial intelligence models and techniques that can generate accurate genetic information, which can be utilized in cloning and selective breeding methods.
Out of the hundreds of thousands of species known to be extinct, only around 8-9% have a DNA fragment that allows cloning. This percentage could go up by 30% if advanced AI models are used to reconstruct the DNA fragments. AI models like Evo 2 use learning patterns from living relatives of extinct animals to guess and fill broken DNA sequences, which traditional methods cannot fully restore. For instance, Colossal – a biotech company – successfully created extinct species of dire wolf pups by reconstructing the DNA fragments and altering the genes of close relatives of them, like the grey wolf, to bring back the hybrid that resembles the extinct dire wolf pups. Furthermore, it helps scientists to put short fragments of DNA together by analyzing overlaps and patterns in data. It can quickly find the right order and correct errors in these fragments, making the assembly process quicker with high accuracy and precision. For example, hifiasm – an AI DNA assembler model- was used by scientists to assemble the DNA fragments of passenger pigeons taken from the museum that assisted in the ongoing research of de-extinction of these birds.
Additionally, AI helps in seeing what ancient species were like by making realistic pictures and 3D representations of them. AI can analyze fossil records, incomplete skeletons, and living relatives to produce visuals of animals that are no longer alive, such as the woolly mammoth, dodo bird, and saber-toothed cat. Mid-journey is one such AI tool that helps in reconstructing the visuals of extinct species by using machine learning algorithms. Besides this, AI aids in organising and storing data of extinct species by creating digital repositories which can be used by others to know about the species and background of the extinct species. For example, Macaulay Library by the Cornell Laboratory of Ornithology uses digital repositories to store thousands of bird species, which also includes extinct species like the ivory-billed woodpecker and Kauai O’o.
Despite these applications of AI in reviving De-extinction, some major challenges need to be resolved. First and foremost, AI relies heavily on DNA sequences to reconstruct extinct species. However, many times, DNA fragments are fully damaged or have missing pieces, which makes it difficult for AI to rebuild an accurate genome. For instance, Dinosaurs cannot be revived using AI and bioengineering methods, as they are known to have been extinct about 66 million years ago, and the survival time of DNA fragments is about a few million years. Secondly, even if AI helps in recreating extinct species, the environment in which the species survived has already changed since it went extinct.
Thus, the relationships between habitat, ecosystem, and climate change will affect the survival of species, which can lead to the failure of reviving many species. Take the example of woolly mammoths that have been known to be extinct for about 10 thousand years ago, the ecosystem in which they lived was grasslands and tundra, which has been replaced by forests and landscape. As a result, this shift in habitat will lead to a struggle in the survival of woolly mammoths, even if they are reconstructed using DNA fragments. Besides this, AI cannot address moral questions like the impact of reviving these species on the ecological system or the welfare of these engineered animals.
Despite these challenges, the future of the De-extinction field, with the integration of Artificial Intelligence, is being shaped by higher accuracy than the traditional approaches and an increase in efficiency. Now, it is no longer just a dream of seeing extinct species again, but it is becoming a reality because of technological advancement. Scientists are trying their best to revive several species, including Tasmanian Tigers – an extinct species of the Australian mainland, 1936 – by assembling DNA using AI. Furthermore, creating more advanced models of AI can increase the accuracy and efficiency of de-extinction research. Alongside, AI can be used to simulate the ecosystem of the extinct species and analyze behavioral outcomes that scientists can use to predict and mitigate risks before actual release.
It would not be an exaggeration to say that AI has transformed the De-extinction field, making the dream scenarios of viewing extinct animals possible. This collaboration not only aids in reviving the species but also restores the habitat and ecosystem of the animal in which it used to live and learning from past experiences to safeguard the current endangered species. Despite major challenges in ethical considerations and accuracy in genome prediction, the rapid advancements in the field of AI will decrease these challenges in the long run.
Furthermore, the combination of genetic insights with the analytical power of AI can assist scientists in restoring biodiversity, preserving the ecosystem, and developing innovative solutions for environmental sustainability. This collaboration of Biology (Genetics) and Computer Science (AI) will foster a deep connection between humans and nature, where technology will aid us in strengthening our bond with the flora and fauna of Earth, leading us towards a better future for society and the environment overall.
(The writer is the First B.Tech Student at Plaksha University, Mohali).
