longstaff2009

Thats a spicy example dataset!

I like that it's able to infer information from the context of the cells, e.g. being able to run a query across continents when the data only contains the country.

Being able to ask it to interpret the results is helpful, it would be cool if it automatically told you if there was enough data to have statistical significance in the conclusions it was presenting.

they have a live demo https://demo.buxfer.com/dashboard

providing end to end provenance of live exports in cross border supply chains involving China.

I wrote a post with my predictions here https://medium.com/swlh/9-predictions-for-2020-2029-555312fc...

TLDR 1. Federated Learning will unlock value from previously inaccessible sensitive data.

2. Deepfakes will impact democracy and bring about a need for publisher certified content.

3. Nationalism will rise around the world, the internet will splinter.

4. eSports will take a huge chunk of attention and advertising dollars from sports.

5. Blockchain will get adoption in enterprise. Mainstream adoption will struggle until there is a key custody solutions that everyone can use.

6. Self driving cars will open up new business models. Regulations will be the main barrier to adoption.

7. Welfare systems will get strained.

8. Digital currencies and negative interest rates will open pandoras box.

9. Search will get reimagined.

I also enjoyed these posts

https://avc.com/2020/01/what-will-happen-in-the-2020s/

https://continuations.com/post/189997132065/an-agenda-for-th...

https://alexdanco.com/2019/12/17/ten-predictions-for-the-202...

in private consortiums, instead of having data locked up in silos across organisations, everyone can have a copy of the data and build their own view on top of it.

you don't have to trust any party as to the authenticity of the data as you can see every change that has been made.

I wrote about the supply chain use case here https://medium.com/twohands/lobsters-on-the-blockchain-df599...

and how I approach looking at distributed ledger tech here https://medium.com/@ben_longstaff/my-framework-for-how-to-lo...

:)

My understanding is that Trump's trade war was apart of the America First economic policy to reduce the US trade deficit. This had the added benefit of making manufacturing more viable in the US.

I found the thought process of trying to make the predictions quite rewarding to go through, it will be interesting to look back in 10 years and see how close or not they are

You are right that section could be much clearer.

Long term transaction data and browsing history could be used to train a federated learning algorithm to generate search results.

Getting access to that data to understand what stories it can tell is difficult ask without offering a lot of value in return.

In the short term I believe providing insights into a persons finances is the way to provide that value.

E.g. if you have an extra $5k how do your circumstances change if you - pay off student debt - pay extra on your mortgage - buy solar - contribute to your 401k / retirement savings

or showing how changes in spending alters the amount of pretax money you need to earn to support your lifestyle.

or how does your spending compare to people like you?

bear in mind that back in 2000 git wouldn't be released for another 5 years. The processes of building software has improved a lot since then

it's an easy trap to fall into, not enough people talk about it

As the author I probably can't change your mind without revealing his real identity, but i'm glad you thought the writing was well calibrated :) so thanks I guess?

I suspect that Dave's pay was not proportionate to his skills

I would say most small engineering teams have that risk

I had missed these, thanks for sharing

Not being a holder of a unique resource is a good point.

Google do have the mind share though, can you imagine trying to explain to Aunty Robyn that you have to type duckduckgo.com into the search bar now? and that she cant google it anymore?

I like this phrase a lot :)

it was, and it sent a strong message to other big companies that Google doesn't care who you are. This is Google's playground and if you want to stay in the search index then you play by Google's rules.

For a token fine I would agree. If the scale of the fines where to increase from a level that can be written off as the cost of doing business to something more damaging, I think that could change things. Particularly as we are seeing a rise in nationalist political movements around the world.

Maybe, but if the French government where to send the army into an American companies datacenter, would America stand by? Particularly if tech lobbyists had been contributing heavily to political campaign donations?

That would be a big gamble for each member country to make.

You are right, it should say "act in the interest of" rather than "make money".

Yes ! At some point the cost of the reputation hit becomes less than the value of the fine.

Google removed 'Don't be evil' from its code of conduct last year. Hopefully not a sign of things to come

I would call it a thought experiment rather than a power fantasy, but sure.

Would it be a thread to US national security in the same way it could be to France? It might depend how the President at the time is and where their campaign funding came from.

Rep. Ocasio-Cortez Proves Just How Easily Corporate PAC Money Can Influence Politics

https://www.facebook.com/NowThisPolitics/videos/vb.908009612...

So I think that is why the fine is 0.05% of global revenue instead of the full 4% that can be enforced under GDPR. Both the government and Google can spin it as a win.

But say instead France asks for 4% of global revenue, and there are rumblings from Spain, Italy, Germany, UK, Sweden, Poland, Romania, Denmark and Greece that if France is successful they will do the same.

Suddenly the potential fines are 40% of global revenue, but Europe doesn't generate 40% of global revenue.

maybe, but when it comes to calling bluffs it would be much harder for the government to sustain continued denial of services than for Google et al to keep the case in court

maybe for some cases, but then there are the government and industry bodies that have exabytes of data stored. if Amazon wont provide access to snowmobile to you it becomes problematic

in the hypothetical scenario where France tries to enforce a huge fine and Google says no thanks, France pushes and then Google turns off service it would be crippling. Google has a lot of power in that sort of negotiation.

Which is why I think it was only a token 0.05% of global revenue for the fine rather than 4%

On what grounds? They would just be changing their business model

I think the extent of that leverage is not well understood.

It's like AMI trying to blackmail Jeff Bezos, but having all of their sites hosted with AWS ....

What happens when it makes business sense for Google, Amazon, Apple and Facebook to start charging nation states a license fee for offering their services into their country and all the government services rely on their cloud offerings?

I did a few migration projects when I worked at the Discovery Channel and they took months of planning to ensure continuation of services.