ivorbuk

Of course it does! Why pay all those machine learning/data employee salaries?

In 2016 I joined an online casino (as a PM) and was confronted by the biggest machine learning and real-time data operation I had ever seen at the time.

All about pushing wagering bonuses, measured by deposits.

At the time we still called this class of user "high value". Not allowed to do that anymore... Gives the game away!

The noise too.

Your "typical" performance car customer wants a monster noise from the engine. A visceral sound. If it sounds like a jet fighter taking off, even better. There's no "authentic" sound to replace it.

Music to the ears (Grand Prix 1966) -> https://youtu.be/5RILdsjeL_4

This is incompetence, pure and simple.

They had incredibly fractured internal systems, multiple owners who refused to invest for the long term and then Tata - who outsourced it all to teams who aren't doing anything with any urgency.

Ticking bomb waiting to happen.

+1 on the demographic base - it's a shrinking market and was geared towards something "quintessentially British"... which only really sells well in the UK. Close your eyes and picture a Jag driver and you see Nigel Farage. Not a fashionable sell.

Pretty proud of their first pitch - it was a brave move. Looking forward to seeing how they get on.

Agree - the Land Rover brand has such significant demand that there's zero risk they won't bounce back.

The "bailout" says more about Tata's approach to supplier relations than it does about brand risk. Not that Tata is much different than any other abusive conglomerate.

This is akin to being "forced" to do the right thing by the UK government. Do we really believe that Tata Group couldn't afford this?

I don't play slots, but I did manage an online casino for 3 years. Auto-spins allow users to play without "interaction" but, at least for my players, the animations and chirps were all part of the "fun".

All go towards the same thing - dopamine addiction as an activity.

I think I just wanted to say that the perception of a 90% "success rate" v's all or nothing with sports betting felt like a troublesome comparison, for me at least. These games arent safer or fairer at all - they are both doing the same thing - in different doses to different audiences.

Many of my slots players didn't consider what they did "gambling" at all.

A 90% RTP will typically (typically - lots of exceptions and caveats) be quite volatile - offering big wins but long losing streaks. Exactly the kinds of game behavior designed to extract long playing times and really peak the dopamine response.

The casino will be accounting/forecasting a 5% take across stakes made as it's super complicated with jackpot contributions/payouts and stuff.

These machines are there to be always accessible and to train the human in just the right way so as to get you addicted.

Judging by the comments on the exploit, it's down to some sloppy estate management - which is typical TCS, sadly.

I know their systems and I can well imagine there's conversation about a greenfield rebuild for the essential manufacturing systems happening or taking place. I can't imagine some of the old processes still being viable after this shocker.

We use llama models as the comparative cost, at the time of decision making (9 months ago?), was cheaper++ then closed source models and comparative to other models available through bedrock (we use via aws bedrock).

Strong reasoning at the time was also to, at some point, take this on-prem/self-host - for privacy. More a comfort blanket for some of our customers/partners and a future requirement rather than a right now thing.

From a capability perspective it's everything we need - tho we are not taxing or pushing any boundaries... Our use cases are mainly processing/structuring/summarising incoming text and we run a few agents doing a variety of stuff. We have a bit of technical jargon (motorsports engineering related) and we saw good performance with llama over our previous use of Mistral/Claude/OpenAI.

S3 buckets for the SPA's, AWS Lambda, API Gateway and EC2 for a MySQL instance. Some SQS to help with orchestration of the operational stuff. Have about 5k MAU and total cost barely breaks 40$ a month.

Chose this because it's the stack I'm super familiar with.

Guestbooks, where people left nice comments.

Getting upset with people for disturbing your concentration whilst having an IM open is like moving to a house by the airport and complaining about the aircraft noise.

There are great tools on these things that help you prioritize and snooze notifications and ensure yourself some quiet time.

My Grandma was already blind by the time the internet became a "thing" and, knowing I worked "on the internet", she would ask me what it was all about when she heard it in the news. After a few struggles we settled on "it's like ceefax, but much much bigger". She knew about ceefax, she was happy she understood what I did. It was a good analogy in the early days (mostly WWW work).

I did get the reputation with her friends that I was a "ceefax expert" and would always be drafted in whenever her friends were struggling. I was "Mr Teletext" for a couple of years. (the other ceefax like service available in the UK)

Another memory was watching the live football scores on a Saturday afternoon. Waiting patiently to get to the right page and the agony when the page you wanted was skipped... having to wait another 8 pages before you were back to the right division (only to find out you'd gone a goal down). Happy days.

About a year into my first IT role we installed a lot of networking infra into schools. Every school, bar none, had their network infrastructure named after characters from Blake's 7 (https://en.m.wikipedia.org/wiki/Blake%27s_7)

The domain controller would inevitably be named Orac. So much so that we set them as defaults in the routers we installed.

A generational thing.

486-dx2-66 in 1995, with CD-rom, maybe 16mb of RAM, 100mb HDD. (maybe 20, maybe 200, it was a long time ago).

Without an OS, self built, cost me approx ~£1600 (GBP). Saved all my McJob money for an entire year.

I remember the cost as it was 100% of the money I saved and my desire for more cpu power meant I couldnt upgrade my crt (amber monochrome).

I brought it all home on the bus.

Edit. 95, not 96.

Duddo (n.) The most deformed potato in any given collection of potatoes.

I use that word to this day.

I think there's something in the names of the families and characters that, at least to my younger mind, helped blend some of the magic into a dream-like state of never ending wandering.

I've made a similar transition. For clarity, I was an engineer for 15+ years before making the switch. I've been a Product person now for about ~6.

As another comment mentioned - I find that PM/TPM roles vary depending on the company and the expectations of your colleagues. It's only sometimes visible from the outside or the job description. This is a big point for me.

I joined a large company (FAANG-like) as a TPM - the interview was great, and I got to talk about technical architecture and solutions. My interview task was to do some 5W analysis on data, dig out the bug, sight reading architecture/code, and talk about iterative product processes as applied to long-term platform work. It seemed perfect. I joined only to find the teams I worked with were amazed I could follow technical discussion/planning and near-offended that I could participate in their conversations.

I joined another large company (again, not FAANG but FAANG-like) as a non-technical PM at the opposite end of the spectrum, UX interviews and labs etc. They loved that I could articulate what I saw more technically and appreciated my experience, involving me in high-level planning and architecture discussions. I ended up contributing to code, architectural decisions and more.

On paper, I was perfect for the TPM role, but it was an excruciating experience that knocked my confidence and depressed me. I believed I was NOT the right person for the out-and-out PM role, but it turned out to be one of the best experiences of my professional life.

I guess my two pennies' worth is to try and understand what the role is going to be, ask probing questions of the recruiter or hiring manager, ask if you're able to meet the team or the EM's you'll be working with - find out how the work is done. You can't change who you are on paper. People will look at your Linkedin profile and always read it their way. I've been the hiring manager for many PM and TPM roles - you'd "stand out" to me by making the right kind of inquiries about how the work is done, how you are expected to contribute, and talking about that.

Worth noting after six years as a PM, I'm considering going back to code. I love being a PM, but code was my first love, and nothing compares to building things yourself.

100% it's this.

I've been in the room when making this exact kind of product proposal (not MB). The pushback from drivetrain and electrical on warranty, repair, battery life etc was intense and directly related to future satisfaction. Sure, you can push the envelope on the spec, but it does tend to come at a cost.

Say they give this option for free... And it reduces the battery life/capacity by 10% over the course of a year. Pretty sure people would be clutching their handbags in horror.

You buy a car with a spec. Aftermarket mods are definately a thing to increase that spec, and they've been around for a while, both for ev and ice. They've always been available at a cost. This time it's from the MFG and we all lose our minds.

I thought I'd never be able to code again - here's my story. There are no revelations here, perhaps just know that people do, eventually, recover.

I spent 6 years at a startup before realizing it was never going to be successful and I walked away (I was the founding engineer and mostly handled the PM work, it felt like walking away from significant other). I was doing the same, crazy 70-100 hour weeks sucking up all my energy. I had no energy or doggedness to get to the bottom of even the simplest issue. That was 2012. I'd been an engineer for 15 years, and I figured I was now too old to keep up. I gave up.

Since then I've been working Product/delivery roles, which I enjoy very much, but there are frustrations (no EM likes a technical PM looking over their shoulder). I tried to code in my spare time on things but the doggedness to learn/fix complex stuff just wasn't there anymore. Also life gets in the way. There are always more important things to do when you've got kids than spend time noodling with code.

Fast forward 7 years I was trying to get a game running for my son, an old 90's Win95 title. I started noodling with Docker and python. Very simple scripts. Suddenly, one night, I found it was still there, all that excitement about code, all that interest and inquisitiveness that had previously been hidden/suppressed came out bursting out. Before long I was pouring code out like hot coffee.

I'm still working my confidence back up to a place where I feel I could interview for a JOB doing it - but for putting time into my side-projects and noodling with Apps, I have all the energy and focus in the world.

TL;DR - I burnt out and changed roles, continuing to support engineering teams. A few years ago I did some code and concentrated on learning, not delivering, and that seemed to unblock me.