7 technologies to watch in 2024 according to Nature
Posted: Thu Dec 12, 2024 9:29 am
According to Nature International Journal of Science, one of the most authoritative scientific journals in the world, advances in artificial intelligence are at the heart of many of the most exciting areas of technological innovation this year. But not only that. There is everything from protein engineering to 3D printing and deepfake detection. More details below.
Deep learning for the synthesis of new proteins
Two decades ago, David Baker of the University of Washington in Seattle and his colleagues made a landmark discovery: They used computational tools to create an entirely new protein.
Today, progress is reduced to the emergence of increasingly SMS Gateway Chile powerful protein data sets, and it is important to correctly “decode” their patterns. Proteins are to some extent similar to documents. Therefore, the sequence-based strategies that underlie large language models (LLMs) and are used in the same ChatGPT can be useful for protein synthesis.
By viewing protein sequences as documents containing polypeptide “words,” LLM algorithms can recognize patterns underlying the architectural script of real proteins.
And in 2023, there was also notable progress in using this algorithm to generate proteins. Some of them use diffusion models, which also underlie image generation tools such as DALL-E. These algorithms are initially trained to remove computer-generated noise from a large number of real-world structures.
Why is this important? Today, protein design has become a practical tool for creating enzymes and other proteins. These proteins, in turn, are used in the development of protein vaccines and drug delivery devices in humans.
Deepfakes detection
The rapid development of publicly available generative artificial intelligence algorithms has simplified the creation of convincing fake images, audio, and video. The fight against generated disinformation will continue for many years to come. And the recent high-profile case of the Taylor Swift deepfake is proof of that.
Deep learning for the synthesis of new proteins
Two decades ago, David Baker of the University of Washington in Seattle and his colleagues made a landmark discovery: They used computational tools to create an entirely new protein.
Today, progress is reduced to the emergence of increasingly SMS Gateway Chile powerful protein data sets, and it is important to correctly “decode” their patterns. Proteins are to some extent similar to documents. Therefore, the sequence-based strategies that underlie large language models (LLMs) and are used in the same ChatGPT can be useful for protein synthesis.
By viewing protein sequences as documents containing polypeptide “words,” LLM algorithms can recognize patterns underlying the architectural script of real proteins.
And in 2023, there was also notable progress in using this algorithm to generate proteins. Some of them use diffusion models, which also underlie image generation tools such as DALL-E. These algorithms are initially trained to remove computer-generated noise from a large number of real-world structures.
Why is this important? Today, protein design has become a practical tool for creating enzymes and other proteins. These proteins, in turn, are used in the development of protein vaccines and drug delivery devices in humans.
Deepfakes detection
The rapid development of publicly available generative artificial intelligence algorithms has simplified the creation of convincing fake images, audio, and video. The fight against generated disinformation will continue for many years to come. And the recent high-profile case of the Taylor Swift deepfake is proof of that.