And we're back! After a brief break we are now right back with regular AI updates and the global challenges they're connected to. For the new readers; welcome to Nural's newsletter! For the veterans amongst you, thank you for your continued support!
This is a particularly interesting edition with several major updates
Packed inside we have
- GitHub and OpenAI have developed a tool that autocompletes code as you write it
- China have developed an AI model that can do everything from write poems to generate realistic images in both English and Chinese
- Voice actors aren't happy that AI trained on their voices are being used to replace them
- and more...
If you would like to support our continued work from £1 then click here!
Key Recent Developments
Demystifying the Draft EU Artificial Intelligence Act
What: AI regulation is a hot topic that will have a huge impact on the way machine learning is carried out and how companies can interact with our data. The EU previously released one of the first documents outlining how they will go about regulating this emergent technology but according to this paper and twitter thread there are some large shortcomings.
Key Takeaway: Make sure to check out the above twitter thread for a succinct summary of the EU proposed regulation alongside an analysis of where it falls short.
GitHub and OpenAI launch a new AI tool that generates its own code
What: GitHub and OpenAI have launched a technical preview of a new AI tool called Copilot, which lives inside the Visual Studio Code editor and autocompletes code as your write it.
Key Takeaway: Those of any level of coding will attest to the utility of autocomplete while coding. The tool would save hours of mindless searching on stackoverflow for code examples and potentially drop the barrier to entry for writing high quality code. However, some ethical questions over code ownership and the data the tool was trained on must also be considered...
Meet Wu Dao 2.0, the Chinese AI model making the West sweat
What: A new artificial intelligence model developed by Chinese researchers dubbed Wu Dao 2.0, is able to understand everything people say — the grammar too — but can also recognize images and generate realistic pictures based on descriptions. It can also write essays and poems in traditional Chinese, as well as predict the 3D structures of proteins.
Key Takeaway: There is a huge rush for dominance in the AI world between different regions globally and this model is definitely a big step for China. What do you think the implications are of the release of this model?
Delhi Traffic? No Problem. Now, AI Will Show You The Way
Other interesting reads
Cool companies I have come across this week
Ro - Ro’s vertically integrated primary care platform powers a personalized, end-to-end healthcare experience from diagnosis, to delivery of medication, to ongoing care.
NoTraffic - Autonomous traffic management platform, reducing congestion and increasing safety.
Just very cool
gpt3-demo - Platform showcasing the companies which are built upon OpenAI's GPT3 language model.
AI/ML must knows
Few shot learning - Supervised learning using only a small dataset to master the task.
Transfer Learning - Reusing parts or all of a model designed 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!