HN user

lucraft

864 karma

danlucraft.com

[ my public key: https://keybase.io/danlucraft; my proof: https://keybase.io/danlucraft/sigs/zV0A03q6S1M9d_zCuBmD_dbnCy5TJQcagAjjzEmwHLo ]

Posts5
Comments178
View on HN

It's a coordination problem. I can "parent" my kids by banning them from social media sure, but if the other parents don't also do that, then I just made my kid a social outcast.

As it is we have to allow our eldest a lot more screen and social media time than we think is healthy, but it's more healthy than not having any friends.

I'm not necessarily in favour of a government ban, but I do wish more parents were on board. At the primary school (age 10) 100% of other kids had phones, and no one else seemed to give a shit.

Right - don’t make subagents for the different roles, make them to manage context for token heavy tasks.

A backend developer subagent is going to do the job ok, but then the supervisor agent will be missing useful context about what’s been done and will go off the rails.

The ideal sub agent is one that can take a simple question, use up massive amounts of tokens answering it, and then return a simple answer, dropping all those intermediate tokens as unnecessary.

Documentation Search is a good one - does X library have a Y function - the subagent can search the web, read doc MCPs, and then return a simple answer without the supervisor needing to be polluted with all the context

To an extent I think would be surprising to anyone who hasn’t done interviews, all developer CVs below about Staff level are the same.

Everyone’s architected a system, handled communication with stakeholders, contributed to technical direction of the team, mentored other developers, etc etc

I tried the demo, fingerprint.com, in:

  - Safari
  - Safari private mode
  - Chrome private mode
and it was not able to identify me across those.

I then tried

  - Chrome (normal, non-private mode)
and it did identify that as a repeat Chrome visit.

Does Safari have better privacy than Chrome?

Plural AI | www.plural.ai | Senior developer | Remote (UK)

Plural creates AI workflow tools for investment banking and private equity analysts to supercharge their productivity and their ability to derive insights on their deals and data. We work with customers in London, Paris, Berlin, and New York.

We're looking for a senior full stack engineer with experience building complex UIs and data handling code on the backend in Python. You'll work on our RAG system and document processing pipelines, so any experience in those areas would be helpful too.

We are a small startup, with a committed team and lots of opportunities to learn on the job, and be given ownership of significant projects.

Technologies: Python, Typescript, React Salary: 80-110 GBP depending on experience

Email dan at plural.ai if interested

Am I right that this article nowhere compares against the current red or even has anything with the current red in it, anywhere?

Is that a policy choice there? They don't want to.. not be future looking or something? I don't really get why. It made the article frustrating to read

Removing nodes by adding a concept of implicit branches via "unless" clauses within a node doesn't necessarily make the system any easier to understand at a glance.

The original diagram does this too - it has an "A && B" node.

Thanks... I haven't thought about it in a long time, as it was difficult, but looking back it seems quite positive.

Three things:

1. Full disclosure at that time in my life I was rather bad at motivating myself to work independently for long periods. In hindsight starting a PhD was a bad idea for this reason alone.

2. The university closed the department I was in. I was transferred to another supervisor in another department, who was nice but saw his role as more administrative. After a while he then announced he was retiring so I was looking at moving to another supervisor again.

3. I turned out to be far, far more interested in writing software than doing research. E.g I wrote an open source unit testing library in Prolog to support my research tooling. I was learning Rails on the side. I went to the Hacker News meet-up in London, and the startup that was running them offered me a job, and the rest is history!

I had sunk multiple years into it so it wasn't easy. But in hindsight it was not even a close decision.

"You are formally invited to A WAKE for THE RESEARCH SCIENCE CAREER of FRANCES HOCUTT FRIDAY from 7 PM to MIDNIGHT"

When I quit my PhD I had an Ungraduation Party! My wife made a cake and everyone sang Happy Ungraduation To You! It was sad and happy but overwhelmingly such a relief to get out

One reason was to increase investment. In water's case the infrastructure had been under-invested in under public ownership for decades, due to it always being a low priority from a political point of view.

The regulatory structure of the privatized industry says the water companies can only make profits if they make capital investments, and so the amount invested in water infrastructure in the UK (particularly England) has duly gone up a lot since privatization and has been comparable or higher to other nations since.

From this point of view, it's less about choice and markets as a way to structure the regulation and ownership so that capital investment actually happens.

At the time I calculated that if all the money spent on the 2012 Olympics (supposedly to inspire a nation's children to love sport) was instead given as endowments, then every single secondary school in England would receive something like £150,000 a year extra in their budget to spend on sport - in perpetuity.