burntoutgray

The managers don't care either. They are job hopping too.

I won't be surprised if companies with large amounts of vibe coded systems end up blowing up.

Why not go the other way around? For example, with Pharo, there is a WebBrowser project which could be further refined to do more of what you are proposing.

Without reference to your own level of knowledge it is impossible to suggest prompts.

What works for me is to pick an area that looks interesting (to me), paste some code and ask questions. Then you drill down until you sufficiently grok that area, then pick another, etc.

> The problem with retirement planning is that you don't know how long you're planning for.

Live it up now! When you are old, frail, bad eyesight, little stamina -- you don't need much. Just a comfortable environment.

... and stop doing things that you don't like just because others are guilting you to do them. I once heard somebody use the expression "x is shoulding all over you!" Really struck home how people try to dictate how you should act/live/think.

A simple advice is to consider the Taoist symbol of Yin-Yang. Move away from the dark and towards the light. Live in your own light!

There's a lot more to life than building startups. Since in your assessment you've done your job, why not take a long break? Review the past 2 years experiences and determine where your passion now lies. With experiences our priorities change. That is normal part of personal development.

I dislike the bloat that Electron apps are burdened with.

Unless you have speed to market, MVP, etc considerations, then why not use the tools that you enjoy using the most?

I could write a boring long comment on why you should do a PhD, but only on a topic that you are really REALLY interested in. Otherwise it is wasted time and missed earnings.

Everybody is different, so it's hard to give advice when you don't know a person's priorities.

If you have the means to focus on getting a masters in a speciality field that you have a strong inclination for, then just do it.

Earning good money doing something you dislike is a golden trap. For most people, a high income leads to an expensive lifestyle. Once you get used to that lifestyle it is near impossible to cut back and follow your passion later.

When you were looking for internships, did you not find any ML/robotics ops? I would have thought that a startup and doing a masters part-time would be a good option. Just need to hustle a bit more. Research potential startups, etc.

The article is dissing ink jet printers.

I have used HP, Lexmark and Brother laser jet printers and they all worked well. I did, rather naively, buy a Canon inkjet to print to print in colour -- a very big mistake. With a B&W laser I can print 99% of my needs and just pop down to the local big box office shop to print the rare colour print and photos.

Not before there is a vast amount of ML training on the facts, experimental results and theories. LLMs trained by scraping the internet have very little "understanding" of hard, real life facts.

As an example, AI in the field of radiology has required vast amount of guided training on radiology images before it was able to provide analysis capabilities. Yet once the need for surgery has been established only trained surgeons are able to perform the necessary operations.

The more experience a developer has, the more they know myriad ways that things don't work out as the requirements spec fails to elaborate.

I've been in this industry for decades. In summary these are my criteria:

* If the solution is software only, then there's probably a 100+ comparable solutions out there. See the point below about competition. I prefer problems that involve real people, real money and tangible solutions. YMMV

* First and foremost: Talk to real people who might have the problem you are solving. Emails, forums, etc are a waste of time. If you can't in person contact with 10 people who have the problem you are solving, then how are you going to scale to thousands of millions?

* How big is the pain you are solving? Getting people to part with money causes them existential pain. So your solution needs to transcend that.

* Check out the competition. If there is none, there probably isn't a market either. If you have competitors seek out how you can deliver 10x value compared to them. If you can't, then move on.

* Can you deliver a demonstrable, working MVP within 6 months. If not, then others will probably beat you in that race.

AI is a tool! Like any tool it is great at what it has been designed to do and not so good when attempting to use it otherwise. At the core LLMs have hoovered up the internet and made lots of statistical inferences. As a research assistant it is phenomenal. As a rational thinker not so good.

A buzz saw is great for cutting lumber; not so good for manicures.

Linux is the kernel. The rest of distros are a collection of bits and pieces that some folks thought might be a good idea.