Welcome to Nural's newsletter where you will find a compilation of articles, news and cool companies, all focusing on how AI is being used to tackle global grand challenges.
Our aim is to make sure that you are always up to date with the most important developments in this fast-moving field.
Packed inside we have
- Alphabet (Google's parent) applying DeepMind breakthroughs to launch drug discovery company
- Meta/Facebook deletes facial recognition data of 1 billion people
- and, a robot talking back against abusive language!
If you would like to support our continued work from £1 then click here!
Graham Lane & Marcel Hedman
Key Recent Developments
Alphabet is launching a company that uses AI for drug discovery
What: Alphabet (the umbrella company of Google) has created a new subsidiary called Isomorphic Laboratories (IL) to build on the ground-breaking work done by DeepMind (another Alphabet subsidiary) in predicting the structure of proteins. IL describes itself as a pioneer in the emerging field of “digital biology” with a mission to use AI to accelerate drug discovery. The CEO of DeepMind will also act as the CEO of IL initially.
Key Takeaways: This is an exciting initiative and will doubtless yield new discoveries. Some sceptics, however, assert that drug discovery is not the real bottleneck in getting cures to sick people given the power of big pharma companies. IL will “partner with pharmaceutical and biomedical companies” and it remains to be seen what this will yield.
Initial IL blog post: Introducing Isomorphic Labs
Facebook/Meta to stop using facial recognition, delete data on over 1 billion people
What: Meta (formerly Facebook) will delete the facial recognition templates of more than 1 billion people. These were used to train the DeepFace face recognition software identifying people in images on Facebook. However, the company will not delete the DeepFace system itself and it does not rule out using face recognition in the future for applications such as “people needing to verify their identity, or to prevent fraud and impersonation,"
Key Takeaways: Facebook experienced a range of regulatory problems in regard to the use of face recognition software. On the other hand, Meta is invested in creating the “metaverse” and has a tie-in producing smart glasses with Ray-Ban, so are unlikely to be leaving this area completely.
Facebook blog post: An update on our use of face recognition
Can feminist robots challenge our biases?
What: Many female-persona digital assistants are patient and accommodating in the face of profanity-laden abuse. Researchers in Sweden experimentally investigated whether an ostensibly female robot that fights back against sexist and abusive comments would prove more credible than one which responds in a subservient manner. Amongst would-be computer science students, girls found the argumentative robot more credible, but not boys.
Key Takeaways: The research clearly touched a sore spot with some (male) users of Reddit and similar platforms. It builds on previous work by UNESCO (I’d Blush If I Could) examining how AI voice assistants projected as young women perpetuate harmful gender biases. Recently Women in Robotics has launched a photo competition of female robotics engineers, particularly those making female robots.
Many companies pay bounties for finding software bugs and vulnerabilities. Could this also work for ethical issues?
Explores the idea of “carbon accounting” to make the carbon footprint of AI systems more transparent.
62 leading German scientists express "deep concern about weapon systems that select and attack targets without real human control".
Other interesting reads
Application deadline is 17 December 2021 for 2-week virtual summer school in August 2022
Describes the first steps towards intelligent firefighting robots that can make decisions autonomously.
Discusses predictive analytics; targeted diagnostics; next-generation radiological tools; and, tele-health
Cool companies found this week
AI data labeling
Sama - despite operating in the competitive data labeling space, Sama claims to offer a living wage and responsible employment practices. It has raised $70 million in round B funding.
Conversation as a Service
Feminist robot tackling bias
Robot playing out subservient and argumentative roles ...
AI/ML must knows
Foundation Models - any model trained on broad data at scale that can be fine-tuned to a wide range of downstream tasks. Examples include BERT and GPT-3. (See also Transfer Learning)
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.
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!