# Links/notes on data, ML/AI, and public policy for James Richardson

###### tags: `AWS` `AI/ML` `FAIR` `EU` `UK`

## Context and purpose
Following our chat (311220) about the enterprise/corporate risk-tool start-up that James is involved with. I mentioned a range of initiatives/concepts/terms. These brief notes provide some basis to those ramblings. I have limited understanding of the situation including James' background/understanding. Data is at the heart of these initiatives and approaches.

## Initiatives/concepts/approaches

### AI/ML

Google staff, like [Laurence Moroney](https://twitter.com/lmoroney?lang=en) are pretty good at providing accessible introductions to AI/ML. Here Laurence provides a useful 7 min long introduction to ML.
{%youtube KNAWp2S3w94 %}

There is increasing adoption by enterprise of cloud computing (+/-s), with AWS having the largest market share (followed by Azure). A lot of enterprise use tried and tested technologies e.g. software written in Java and data in relational databases accessed using SQL (my expertise is not about understanding 'enterprise' use of technologies and tools). Andy Jazzy (AWS CEO) would like all enterprise to migrate to the cloud (or at least hybrid infrastructure: mix of on premise and cloud compute and storage). Here is Andy's recent re:Invent key note (3 hrs long), at [1:48:00](https://youtu.be/xZ3k7Fd6_eU?t=6494) you see the three layers of AWS ML capabilities (lowest level: frameworks like Tensorflow; middle: Sagemaker tools; highest level: AI services they offer).

{%youtube xZ3k7Fd6_eU %}

## FAIR data
Connected to ML/AI and the public policy intiatives below is data, and there has been a lot written about the rise of data science and data engineering over the past decade. In an enterprise setting then you can find plenty to read/watch related to the concepts of data warehouse and data lakes, [for example this page from Snowflake](https://www.snowflake.com/trending/data-lake-vs-data-warehouse).

I mentioned the concept of [FAIR data (a set of principles)](https://www.go-fair.org/fair-principles/), at the highest level these principles maybe useful for all data-driven enterprises i.e. data needs to be findable, accessible, interoperable, and be reusable.

## EU twin green and digital transition

Ursula von der Leyen's first State of the Union speech. Specific initiative include the [European Strategy for Data][(https://ec.europa.eu/digital-single-market/en/european-strategy-data) and the ongoing work related to the [EU taxonomy for sustainble activities](https://ec.europa.eu/info/business-economy-euro/banking-and-finance/sustainable-finance/eu-taxonomy-sustainable-activities_en).

{%youtube q8eThxTJxRk %}

## Scotland and wider UK
I mentioned the [Scottish tech ecosystem review](https://www.gov.scot/publications/scottish-technology-ecosystem-review/) carried out by Mark Logan and that the Scottish Government (led by Kate Forbes) has accepted all 34 recommendations. Here is [Kate Forbes (Cabinet Sec for Finance) talking to Gillian Docherty about the AI Strategy for Scotland](https://www.thedatalab.com/podcasts/kate-forbes-and-gillian-docherty-discuss-the-ai-strategy-for-scotland/) (that is due to launched in the next couple of months).

In the past year the UK Government has launched a [National Data Strategy](https://www.gov.uk/government/publications/uk-national-data-strategy/national-data-strategy) and [UK's Geospatial Strategy 2020 to 2025](https://www.gov.uk/government/publications/unlocking-the-power-of-locationthe-uks-geospatial-strategy).