hectormalot

Since there currently is no such world-wide central party, why would a blockchain based solution be successful if the central party is (apparently) not?

In this specific example: are car dealers not participating because they don't trust the central party (case for blockchain) or because (lazy|not in their interest|inconvenient|...)? If the latter, then blockchain is not solving the actual problem.

For me that’s Heroku. I just push my app (RoR) and done. I’ve actually moved it once to k8s for a few months and decided to move it back after I understood how easy heroku made the opa side of the business.

Note: it’s a side project, 50 business users, $5k annual revenue, 4h per month as target for the time spent on customer support, admin and maintenance. So it’s both easy to pay heroku and important for me to not spend too much time on Ops.

I think its possible the targeting works for the advertisers (at some level) while still failing completely for end users. e.g. If I get my hit-rate up from 1% to 2% by moving from contextual to 'give-me-all-your-data'-targeting, then that doubles my return as an advertiser, while as a user I don't see a difference and still think: "all ads I see are terrible".

The lie is telling users that targeted ads are good for them

Versioning: I use git to version my ledger file.

Preventing regressions: Every time I update with the latest transactions I close the last transaction with an assertion that ensures the balances are correct. (E.g. that the balance on checking is equal to the latest statement, or that outstanding VAT is 0 after paying taxes)

I’ve been using ledger for my (small) business for about 5 years now. I’ve a script that calculates all the figures for my quarterly VAT submission by the categories that the tax authority requires (I use tags to label transactions).

Importing transactions is still manual work, but given the low volumes it’s not a problem. I do find it frustrating that most banks are still so limited in providing API access to accounts.

Invoices are a bit special, I can see outstanding amounts under Assets:Receivables, but haven’t found a good way to track this easily by invoice.

I think the crux on the vacuums was that people thought: more powerful is better, but they don’t actually vacuum better. (Physics: the max pressure difference from room to vacuum is about 1 bar. Given a certain gap between the vacuum and the floor, the marginal benefit of a 3kW vacuum over a 1kW one is small)

But buying psychology just promotes inefficient vacuums at that point, so it could make sense for the regulator to try and fix that. (From an environmental perspective I agree with this one)

Note: I’m not even sure if the EU even legislated this at all. Turns out, a lot of so called EU regulations bashed by the UK tabloids were represented overly simplified to fit a desired narrative.

E.g. here is Wikipedia on how both sides used the regulation on food to ridicule and defense the EU:

> This regulation requires that bananas as a minimum standard must not have "abnormal curvature",[5] although no definition or guidance was given about the degree of curvature that would be regarded as "abnormal". This led to various stories about an EU ban on curved bananas.[6] This has been frequently repeated by pro-Europeans and Euro-sceptics alike; the former tending to regard it as an apocryphal or misleading Euromyth[6] and the latter regarding it as an example of needless European bureaucracy.

Link: https://en.m.wikipedia.org/wiki/Commission_Regulation_(EC)_N...

I’ve used Mibo for a bunch of department meetings at it was very well received! One of the great things is that it allows you to have serendipitous meetings with people you haven’t spoken to for a while just by running into them on the island.

Great team as well! They helped us set it up in a way that also worked for the corporate firewall.

I believe they use Jitsi as their backend, combined with some cool engineering to make it all work on k8s.

> The only way EIA will be off the mark is if we get something like fusion. Currently the major geopolitical issue is natural gas pipelines. Major nations wouldn't be fighting over gas pipelines if people believed fossil fuels were over.

I'll point out that the EIA has been off the mark wrt renewables in the past. For example: in 2010 [1], the EIA predicted renewables to grow by 30 Quadrillion BTUs in the next 20 years (2010-2030), that took only 8 years instead of 20. In their latest report they had to update that to a growth of 120 Quadrillion BTOs over the same timespan (2010-2030) (i.e. they were off by a factor of 4x).

As a thought experiment: The world doubled renewables in the last 10 years. If we maintain that CAGR until 2050, that would mean growth from ±100 Quadrillion BTU in 2020 to 200 in 2030, 400 in 2040 and 800 in 2050. If the totals from the EIA in 2050 are correct (±900 Qd BTU), then that would cover almost 90% of primary energy consumption globally.

Note: I realise this is oversimplified. We'd run into dampening factors (e.g. intermittency of solar/wind) far before we reach 90%. Nonetheless, based on the current growth I'm happy to argue that the EIA is underestimating growth potential of renewables.

> Right... One of the biggest per capita producer of oil is norway. > The 2nd largest oil company in the world is a anglo-dutch company.

The present doesn't have to describe the future:

- Shell is still huge company. Yet, had you invested in Shell 5 years ago, your stocks would be 33% lower today (-47% for Exxon). Had you invested in some of the established renewable energy players e.g. Orsted, Vesta, or Iberdrola, you'd be up between 75 and 350%.

- Same in the US: Nextera (US Wind and Solar player) is currently valued higher than BP, Total or Shell. ($147 bln), and doubled market cap in the last 5 years.

The Dutch version of the FT ran an article yesterday on how investors are increasingly wary of investing in fossil fuels, and moving more into renewables. They made the comparison to how it suddenly became undesirable to invest in e.g. Tobacco. [2]

[1]: http://large.stanford.edu/courses/2010/ph240/riley2/docs/EIA...

[2]: https://fd.nl/ondernemen/1366175/beleggers-draaien-van-fossi...

I hope they (EIA) are off the mark. If the predictions from the EIA a will materialize, we’re going to have a huge carbon problem.

However, I think there also are reasons to believe things will be better:

1) the EIA forecasts linear growth of renewables, while its been behaving exponentially over the last 5-10 years. That gives me hope they’ve underestimated the long term growth rate. (It would be interesting to compare their older predictions vs actual)

2) nearly all of the energy consumption growth is non-OECD and will require newly built generation. If cost of renewables continues to drop, it might just be economic sense that pushes renewables over coal

3) at least on the EU side, fossil fuels are increasingly getting a negative connotation. On top of the financial comparison, I expect this could dampen new fossil growth as well.

I agree that many people don’t know what they want. However, I think this is less about emulation and more about improving ones options while figuring that out.

Consulting, Finance, and Bigtech offer a combination of rapid salary growth, low career risk, and a resume item that almost ensures that you pass the CV stage at every future interview. If you’re unsure about your path it is a great way to ensure you have many doors to choose from once you do.

I once read a book that described this line of thinking as the “deferred life plan”.[1]

At some level I think everyone is pushing some part of ‘what I really want to do’ to the future. I know I do. At the same time I think it’s important to remember to enjoy the years leading up to step 4, even if that makes the overall path less efficient.

1: the monk and the riddle

Sorry to hear that. For what it’s worth: I once submitted a tiny pull request for Bazel (IIRC allowing insecure connections to localhost for docker repositories) and it was accepted and merged within a few days. Positive experience overall.

> Want to ban group X which is fighting with group Y in a tit-for-tat skirmish over decades? How do you decide who is right? How do you decide truth?

I think there is a difference between ‘right vs wrong’ and ‘factually correct vs incorrect’.

Platforms like Twitter will rarely be in a position to decide right/wrong, because in many opposed discussion both sides will use the platform.

On the factual correctness, platforms can more easily take a stance in high impact cases. I would assume that they’d have out the same warnings on Biden’s Twitter if he had said ‘I won’ on Tuesday.

That leaves a challenge: how do you decide what’s a fact? For example on the vote count. Most outlets say the reported figures are fact, but it implies trusting the process of counting and reporting. Removing that implied trust means others can take ‘there is fraud!’ as fact just as well.

Twitter seems to use the mainstream media (eg AP predicting the election outcome) as a base of trust (and so do I). But I’ve come to realize that once you don’t trust the media, it is easy to believe in another set of facts (that are true for you), and suddenly (to you) it seems Twitter is censoring.

I hope people will find and trust each other again.

> Everyone on the development team for the project I'm working on is silently in agreement that our model would be better off being replaced by a well-managed rules engine

That was one of the better insights with our team. We should measure the value-add of ML against a baseline that is e.g. a simple rules engine, not against 0. In some cases that looked appealing (‘lots of value by predicting Y better’) it turned out that a simple Excel sort would get us 90-98% of the value starting tomorrow. Investing an ML team for a few weeks/months then only makes sense if the business case on getting from 95% to 98% is big enough in itself. Hint: in many cases it isn’t.

That, and/or mastery learning. I haven’t read the research, but I hypothesize the focus on ‘mastery in each subtopic before moving to the next’ is what drives results, while 1-1 tutoring is what drives the speed of attaining them.

I think gp meant prepayment including all future interest, which would more than cover prepayment fees.

High-lvl: prepayment fees is roughly equal to interest owed minus alternative interest options the bank has (eg someone else’s mortgage). Disclaimer: Dutch perspective.

ASML can sell if they want, but indeed the US has a lot of ways to make that unattractive. Case in point, the US has recently started pressuring the Dutch Government to prevent sales by ASML to China. I think the current stance is that they (Gov + ASML) are still considering how to proceed.

An interesting aspect though: I think the technological superiority of ASML is a big asset in these discussions. e.g. If the US would follow up on a threat of sanctions against ASML (e.g. banning US companies from working with ASML) it would hit the US semiconductor industry quite hard (particularly Intel), while non-US players (e.g. TSMC) could continue producing with euv.

In effect, they would have to go much further to keep some similarity of a level playing field: e.g banning all products containing chips produced on ASML machines. That would hit a lot more companies, many of them US based too. Creating even more reasons for the US Gov to not to take such drastic steps.

Therefore I expect that it is much more likely they'll use pressure on the EU / NL government systems to force the behaviour they want, compared to economic sanctions vs ASML.

> ASML is integrated into the dollar-based global financial system, so it will have to respect American financial sanctions

I'm no expert, but if a company wants to pay ASML in Euros, I'm pretty sure they can accept euros. The dollar is convenient, not a requirement.

I think GitLab does a decent job of that. I think There is a list of countries and how you could work with them. I.e. employed vs contracted vs not-at-all

From a legal perspective I get the limitations to local countries. I considered hiring remote (currently not easily possible due to company policies on eg in presence card pickup and activation) and would’ve scoped to the EU only in that case.

There are great people available outside of the EU, but the additional legal hassle makes the trade-off not worth it unless you expect a large amount of remote employees in that country.

Agree. We do something similar:

* 1 month (or less) to prove that it makes sense to continue. We do start out with a pretty well defined problem definition though. Outcome is typically a bunch of jupyter notebooks proving that we get at least some predictive value that helps solve the problem. If no proof: no-go

* 2 months to upgrade it to something that works well enough to run a real pilot (e.g. shadow run, run with a few agents, in an A/B test, etc.). If no success: no-go

* Then hopefully in production within another 3 months. We're in a pretty regulated industry, most of this time is _not_ engineering.

The phases help set priorities as well. Doesn't always make sense to spend (much) more time on scalability if we don't know it works yet.

Not GP, but I’m assuming https://www.passwordstore.org/. Pass is a CLI password manager using Git and GPG to store and manage your passwords. I use it and sync passwords across devices through a git repo on self-hosted Gitea. Works really well if you’re comfortable with the CLI.

Also switched away from Google to Fastmail with my own domain about a year ago. Happy with the move, although Gmail still sees some activity because I never took the effort of migrating historic mails and all non-primary accounts.

As someone who works in this field: the challenge is more nuanced. For example: Even if you don’t input ethnicity as a value, you might input ZIP codes. ZIP codes however can be very predictive of someone’s ethnic background, and you might still end up unfairly discriminating against certain groups of people.

The Amazon recruiting AI failure is another great example of this, I think it didn’t take gender into account, but discriminated against activities like ‘head of women’s association’ etc.

Finally, what’s ethical is often up for discussion. Depending on context, people will optimize for equality (equal % of credits approved for both genders) vs helping those behind (Higher % approved for those historically disadvantaged) vs helping those that ‘deserve’ it of groups were unknown (finance such that default rates of both groups are expected to be equal, likely benefiting those with already good credit scores). I don’t think any of these answers is wrong, but they are mutually exclusive.

One challenge I don’t see discussed that often is that the current system of many humans evaluating loan applications leads to ‘noise‘ in outcomes so that we (on average) don’t see that much bias. (Eg a few people of certain ethnicity not getting loans because a few people are biased against them)

A single AI system scoring all applications however is consistent in its bias against a group, leading to more obvious effects (eg most people in a certain area not receiving loans because the algorithm is biased against it)

In my opinion, an important part is the right on human review of meaningful automated decisions. Not mentioned that often, but it’s also part of the GDPR.

Off topic, but that reminded me of some horror stories from my chemistry days:

Dimethylmercury able to kill through 2 drops absorbed through a latex glove: https://www.acsh.org/news/2016/06/06/two-drops-of-death-dime...

And of course the amazing series of ‘things I won’t work with’. E.g.: https://blogs.sciencemag.org/pipeline/archives/2014/10/10/th...

ABN AMRO | Amsterdam, Netherlands | Senior AI Specialist | ONSITE

We're hiring! ABN AMRO is a top 3 Dutch bank and we are expanding the AI team within our central innovation department. We're a small (10 people) team and work on some of the toughest AI challenges in the bank. You'll work on creating central AI capabilities and will be able to work with our innovation ventures (internal start-ups).

We are looking for profiles with broad experience across ML, entrepreneurship and business topics. In other words: Generalists. Also come and talk to us if you have experience in building infrastructure for data exploration and deploying models to production. We're using python and tensorflow for models; Kubernetes, jupyter, CI/CD, etc., for deployment

Reach out to the email in my profile for more information.

Note: Please mention HN in the e-mail. We're require a valid work permit for the EU.

The point of futures is that I could sell my oil _today_ to be delivered in June for $20. Any change in June oil prices from that point onwards doesn’t matter once I’ve sold the contract.

> Moreover, you may quickly realize much of this work is repetitive and while time-consuming, is “easy”. In fact, most analyses involve a great deal of time to understand the data, clean it and organize it. You may spend a minimal amount of time doing the “fun” parts that data scientists think of: complex statistics, machine learning and experimentation with tangible results.

This. Universities and online challenges provide clean labeled data, and score on model performance. The real world will provide you... “real data” and score you (hopefully) by impact. Real data work requires much more than modeling. Understanding the data, the business and value you create are important.

As per #6, better data and model infrastructure is crucial in keeping the time spent on these activities manageable, but I do think they’re important parts of the job.

I’ve seen data science teams at other companies working for years on topics that never see production because they only saw modeling as their responsibility. Even the best data and infrastructure in the world won’t help if data scientists do not feel co-responsible for the realization of measurable value for their business.

Training integrative data professionals could be a great opportunity for bootcamps. Universities will (understandably) focus on the academically interesting topic of models, while companies will increasingly realize they need people with skills across the data value chain. I know I would be interested in such profiles. :)

ABN AMRO | Amsterdam, Netherlands | Senior AI Specialist | ONSITE

We're hiring! ABN AMRO is a top 3 Dutch bank and we are expanding the AI team within our central innovation department. We're a small (10 people) team and work on some of the toughest AI challenges in the bank. You'll work on creating central AI capabilities and will be able to work with our innovation ventures (internal start-ups).

We are looking for profiles with broad experience across ML, entrepreneurship and business topics. In other words: Generalists. Also come and talk to us if you have experience in building infrastructure for data exploration and deploying models to production. We're using kubernetes, jupyter, CI/CD for models, etc.

Reach out to the email in my profile for more information.

Note: Please mention HN in the e-mail. We're require a valid work permit for the EU. Mass/generic e-mails will not get a response.

I used the Stavros one: https://www.stavros.io/posts/how-to-configure-wireguard/

Great to see 1.0.0 released. I’ve been using it for a VPN to my home network and have been really impressed with how fast it is (and how fast it connects!). The corporate VPNs I’ve used in the past we’re so much slower that it feels like a completely different league, although they support many more features of course.

> PII may only be processed if you have explicitly consent for the exact purpose you want to use it for.

Not exactly. Consent is one of six allowed ways to process PII. It’s just that for advertising/tracking use its probably the only one that you can use.

No Schengen stamps if you already live in the Schengen area. Unless there is data sharing the US gov wouldn’t be able to see it from a passport.

I suppose it will be another question on the form. Next to ‘I am not a terrorist’

ABN AMRO | Amsterdam, Netherlands | Senior AI Specialist | ONSITE

We're hiring! ABN AMRO is a top 3 Dutch bank and we are expanding the AI team within our central innovation department. We're a small (10 people) team and work on some of the toughest AI challenges in the bank. You'll work on creating central AI capabilities and will be able to work with our innovation ventures (internal start-ups).

We are looking for profiles with broad experience across ML, entrepreneurship and business topics. In other words: Generalists. Come and talk to us if you have experience in building infrastructure for data exploration and deploying models to production. We're using kubernetes, jupyter, CI/CD for models, etc.

Reach out to the email in my profile for more information. You can find more information here: https://careers.abnamro.com/job/Amsterdam-Senior-Artificial-...

Note: Please mention HN in the e-mail. We're require a valid work permit for the EU. Mass/generic e-mails will not get a response.

Crucially: if you’re more likely to deliver successfully you’ll be able to price accordingly as well. E.g. if the perceived change of success with the 60k person is 80% and with you it’s 90% then you’re arguably worth a few 100k extra.

(Now: that %-chance-of-succes is difficult to measure. Which is why it’s a relationship business IMO :)