dasboth

You say you "used the ChatGPT on one screen, code editor on the other approach for a long time" but moved to "prompt and accept".

Can I ask why? Is it work pressures, the convenience of it, or something else? I'm trying to resist making that switch myself but the few times I've used Claude Code for example have made me think whether to use the second method more...

Textual apps powered by sqlite running on a phone sounds like a superpower for filling your phone with useful tools, thanks for sharing! Does it access a local sqlite file or are you hosting that somewhere? Either way, this is really neat.

The Python Challenge sounds similar but I don't know if there was a rock paper scissors component.

http://www.pythonchallenge.com/

SEEKING WORK | UK | Remote preferred | Data scientist | Instructor/trainer

I'm a data scientist and educator with a background in software and web development. Having worked a variety of technical roles, I've realised my two passions are: investigating the value contained in data by rapid prototyping (building proofs-of-concept to iterate ideas), and teaching people the value of having some level of technical skill. In the last couple of years I've been developing and delivering bespoke training programmes in Python and data analysis for enterprise clients (anything from half-day sessions, to weekly bootcamps all the way to 10 week part-time courses).

I'm looking for either a short-term contract (6 months max.) or a longer-term flexible freelance relationship. Remote preferred, but hybrid with occasional on-site (London or Midlands) is a possibility. I'm open to job descriptions beyond "data scientist", since my goal is to actually add value to an organisation through contracting/consulting.

--

Services

· Data science: analysis with the Python data science toolkit, building machine learning models, rapid prototyping to test ideas at an early stage

· Training: development and delivery of bespoke training for both junior/experienced analysts and also non-technical teams. Topics: programming in Python, data science (analysis, visualisation, BI tools), machine learning

Languages/skillset:

Python, SQL, BI tools (Tableau, Power BI, even years of Crystal Reports), and web dev stack (HTML, CSS, JS, C#, ASP.NET MVC)

--

Website: https://davidasboth.com/

LinkedIn: https://www.linkedin.com/in/david-asboth-9256772b/

Agreed, the engineering side is where valuable growth is/will be. Also from the limited sample I know of (people I know who are hiring managers) the candidate pool for data scientists lacks people with genuine analytical ability. Training models is easy, thinking the right way about data is hard.

Chris Albon has a great pack on his website chrisalbon.com

You say "I've enjoyed them much more than the machine learning I've done". I actually find the machine learning the LEAST interesting thing about a data science career in the sense that it's just another tool in your DS toolbox to solve open-ended problems that no one has properly tackled before. It sits in the same box as bar charts, web scraping, asking better questions etc. and you can do valuable work with a variety of approaches, not all of them machine learning. Not sure exactly what part of data science you feel you don't enjoy, and it sounds like you DO enjoy a variety of technical disciplines, but I was just curious whether you wholly equate data science with machine learning.

What about paying for Google services? I do worry about my dependency on the GSuite, perhaps being a paid customer is the way to secure yourself a bit. Looks like the lowest paid tier is £15 a year, I should probably pay that!

I was invited by a friend to contribute to a birthday card on it and loved the experience. Great job, will use this in the future!

This is so spot on about Data Science in the "enterprise" or "legacy" organisations (i.e. basically pre-dating the data hype).

Personally getting stuff done with data in this environment is more satisfying than using the latest neural network, I presume you're the same?

I'll second this. Switching from browsing news sites daily to just reading The Economist (in print, not online) has made me feel no less well-informed, if sometimes a few days out of date, but much less stressed about "keeping on top of the news", which is a losing game.

Interesting - hardcore in what sense?

raptitude.com is a great source of articles to remind you of the things that were compelling to you at any point about mindfulness and/or meditation. I also recommend the spin off http://campcalm.com if you're interested in meditation in particular.

If you're after books specifically, "Wherever you go, there you are" by Jon Kabat-Zinn is also good.

I think the guy's doing that one on purpose to be honest. He's cashing in on the popularity of his (unintentionally) hilarious rendition of Big in Japan (or "bikicsunáj"): https://youtu.be/E2GFjXmXEsw

Python for all the reasons people have mentioned here.

Tangential, but does anyone know of a playground environment where you can code (ro)bots in Python? I'm thinking of the old Robocode challenges, which were Java based, where you could write some logic and it would enter your bot in a tournament. Would be fun to do some game AI with Python like that.

The links on the site are usually the "most trending" ones, so it's good for that, but I wish it had as much discussion as HN. For some reason I much prefer HN to reddit for discussion.

Came here to post this. I swapped my daily news reading habit for an Economist subscription (paper only) and have not looked back. The first 1-2 pages will typically summarise the week's news well enough for me not to feel like I've missed out on anything, and the quality of writing and variety of interesting topics is excellent. I also agree with your second point about feeling obligated to read it!

Edit: for some humorous and UK-specific news/investigative journalism/cynicism I can also highly recommend a subscription to the fortnightly Private Eye. It works out at like £1 per copy and is always a fun read.

Seconded.

AntennaPod supports the two ways I want to add a podcast: search and by url. Haven't run into the problem of it not finding anything either (but if I did, I agree that it's not the app's fault).

Recommend it 100%.

I have a couple of go to pieces on procrastination/getting things done:

Procrastination is not Laziness (http://www.raptitude.com/2011/05/procrastination-is-not-lazi...)

6 Harsh Truths That Will Make You a Better Person (http://www.cracked.com/blog/6-harsh-truths-that-will-make-yo...)

As a jobseeker I use this daily, thanks for a great product. The more data sources you can integrate in the future, the better!

I'm curious, what were the details that made that story sound less bad than it does at first glance?

I find this as well. My motivation for new projects is 100x compared to my motivation to finish existing ones, and I feel part of it at least is because on the existing ones I've reached a point where I'm not learning anything from finishing them. The flipside is that I have a lot less work that I can point to having done, compared to how much I've actually done. Having 5 completed things to show is better (for "achievement" purposes) than 20 projects at 80%, but the rewards as you say favour the latter.

This year I'm thinking about only working on projects where I'd actually use the outcome for something. If it's a project for the sake of learning library X, I won't do it. If it's a project to solve problem Y for myself, I'll try and work on that, learning anything I need along the way.

Location: London

Remote: Yes

Willing to relocate: Possibly within the UK

Technologies: Data science & machine learning, Python data science stack (numpy, pandas, scikit-learn, matplotlib etc.), Javascript, SQL, C#, PHP

Résumé/CV: http://davidasboth.com

Email: hello [at] davidasboth [dot] com

I'm an experienced software developer with an MSc in Data Science looking for my first full-time Data Scientist job.

I bring a blend of programming experience and a new-found love of data science and machine learning, and would like to work on challenging, data-driven projects and keep learning about data science on the job.

That's true. The big if is whether we can ever have the resources to fully simulate the human brain. If we ever get close to that, then we can worry about the "intelligence explosion" and so on. It feels very possible that we'd work out how the human brain works at some point, though.

> I doubt the sun provides enough energy for us to fully simulate the sun.

That's an interesting way to put it! I tend to agree that simulating a human brain will need more than the computational resources to train a neural network with 1 billion neurons (if those resources will ever be available, which is questionable). That's not to say we won't find a way around that problem though. Of course, all the worry about a dangerous AGI hinges on this premise, and how we get there is at best unclear at the moment.

Is it not possible that we could simulate a star though? Not in 2017, but at some point?

If you could encode enough information about the image and have a good training/test set, an AI's ability to determine something as funny probably wouldn't be bad. Anyway, as a subjective concept you might not accept another human's explanation about why something is/isn't funny. This isn't a simple classification problem though, probably suited to something like reinforcement learning.

There is research directly addressing "transfer learning" where the goal is to have more general learning agents. I don't know how advanced it is, but it's certainly not advanced enough to become anything close to an AGI by 2018.

Ultimately I think there are far more pressing issues in AI around ethics, bias, or security. It's great that there are philosophers who can sit around and worry about a possible distant future (and I think as thought exercises they're fascinating), but it's not what most people in the field should be concerned with.

With the success of things like recurrent neural networks for text generation, an entirely computer-generated journal would be interesting in itself.

I've seen similar and can't help but wonder if at least some of those people don't realise that's happening. Maybe they see the "login with Facebook" as a convenience without knowing their names, jobs etc. will be visible. The rest probably just don't care.