Welcome to Nural's newsletter where we explore how AI is being used to tackle global grand challenges.
As always in the newsletter you will find a compilation of articles, news and cool companies all focusing on using AI to tackle global grand challenges.
Packed inside we have
- Mimicking the human mind might be the key to unlocking the next level of AI
- Researchers have discovered a method of detecting deepfakes by comparing reflections in the eye of those in the photo
- Recreating videos of Doctor Who using brain activity from someone watching the show
- and more...
If you would like to support our continued work from £1 then click here!
Key Recent Developments
What: Researchers at MIT have created an ML technique which calculates the binding affinities between drug candidates and their target proteins in the body. The affinity is a measure of the strength of the bind and naturally a stronger bind is desired for the most effective drug action.
Key Takeaway: It should lead to faster drug discovery and protein engineering as the ability to predict the "stickiness" of a bind between the drug and the target protein is a good indicator of a drug’s potential effectiveness.
What: OpenAI have published a review of some of the applications that have been launched using their GPT3 API. Where GPT3 is their large language model with billions of parameters.
Algolia - "Uses GPT-3 in their Algolia Answers product to offer relevant, lightning-fast semantic search for their customers."
Fable Studio - An AI Character Engine, powering story-driven Virtual Beings
They have put out a call for healthcare & climate based applications amongst other areas.
What: Despite the impressive progress of AI systems, they are still no match for the human brain in terms of flexibility and adapting to changing information. Therefore, Nara Logics have launched a platform that mimics the human brain to allow for better understanding of causality and working with unlabelled data.
Key Takeaway: The ability to understand why an ML algorithm makes the recommendation that it does is essential for the increased adoption of AI especially in fields like healthcare.
What: A nice overview of what to consider to effectively use AI to tackle climate change. This is particularly useful for those starting climate related AI projects.
Key Takeaway: Recognising the climate impact of using AI itself is sometimes missed when launching AI climate solutions. A large amount of energy is required to train models and so this must be included in the cost benefit analysis of whether to move ahead with a project.
Other interesting reads
🚀Brain2Pix: Fully convolutional naturalistic video reconstruction from brain activity (Translation: They looked at the brain activity of someone watching Doctor Who and used AI to recreate the images seen!)
Cool companies I have come across this week
Nara Logics - Builds a network of explicit and inferred connections between chaotic, siloed enterprise data for real-time, context relevant recommendations and gives the reasons behind them.
Open Climate Fix - Non-profit research and development lab, focused on reducing greenhouse gas emissions. Their approach is to search for ML (Machine Learning) problems where, if they solve a well-defined ML task, then there is likely to be a large climate impact.
AI/ML must knows
AutoML - The process of automating the process of applying machine learning to real-world problems. AutoML covers the complete pipeline from the raw dataset to the deployable machine learning model.
Transfer Learning - Reusing parts or all of a model desinged for one task on a new task with the aim of reducing training time and improving performance.
Tensorflow/keras/pytorch - Widely used machine learning frameworks
Generative adversarial network - Generative models that create new data instances that resemble your training data. They can be used to generate fake images.
Deep Learning - Deep learning is a form of machine learning based on artificial neural networks.
Nural Research Founder
If this has been interesting, share it with a friend who will find it equally valuable. If you are not already a subscriber, then subscribe here.
If you are enjoying this content and would like to support the work financially then you can amend your plan here from £1/month!