medhir

+100. I’ve found the “chat” interface most productive as I can scope a problem appropriately.

Cursor, Windsurf, etc tend to feel like code vomit that takes more time to sift through than working through code by myself.

you hit the nail on the head. my experience with prompting LLMs is that providing extra context that isn’t explicitly needed leads to “distracted” outputs

I’d really like to bring the fight to the US in a more meaningful way!

can you expand on what you find worse about the concept of having full access to any computer you own?

I’m principled in my stance, I think if you own said Xbox you should be able to load software on that as well.

on the topic of hiring - the “life story” interview left a particularly bad taste in my mouth.

I didn’t click with the recruiter because she kept hounding me about why I didn’t finish college. Felt totally put on the spot in the most discouraging way, the whole thing seemed like a formal way to discriminate against those that don’t fit their in-group.

definitely felt a bit cult-ish overall.

once again will ask why iPhones are treated any differently from other computing devices — we need legislative solutions that allow consumers to load software (or even other operating systems) on any computing device they own.

restricting software distribution on any platform under the guise of needing to be kept “secure” always seemed anticompetitive to me - that should apply regardless of Apple’s particular behavior with the courts in this example.

looking at the way copilot was rolled out at Microsoft, none of this is particularly surprising.

early on, they shoehorned LLMs into most of their products in the least thoughtful way possible — I guess leadership thought it best to move quietly for first-mover advantage.

I remember looking at copilot enterprise pricing at that time and thinking “this is delusional.”

and it still is! $31.50 per user, per month… that is more than most other LLM “pro” subscriptions. the value prop just never made sense when the outputs tend to be worse than what you get using ChatGPT.

not a solution for every single type of JS error, but reading through this I found myself wondering why not just use .then().catch() statements when making async calls.

Compared to try / catch with await, falling back to promises at least makes the error handling explicit for each request — more along the lines of what Go does without having to introduce a new pattern.

omg, this was made at sfpc! spent some time there in 2016, love to see things being made there still.

the crt shader aesthetic is really cool. will try and make a pr of my own sometime :)

I’m using Keycloak as my auth server. Took a bit to figure out the configuration necessary.

Tbh I don’t feel like most providers (including Keycloak) are offering strong, turn-key solutions for this.

The closest I saw to streamlined passkey support that you can host yourself is from Hanko.io - that provider didn’t work for my use case but something to consider.

we have passkeys… unfortunately it doesn’t seem like the narrative really took hold in the mainstream.

I’ve been building an app with passkey auth as the default and people are surprised that such an experience exists.

I sincerely believe these “cute” style airline safety videos need to be regulated away.

Spending so much on production and the content is actively distracting you from the core message… just totally lost the plot.

Personally found the best mileage out of sticking to chat interfaces.

Using Cursor, I felt like it was too easy to modify too many things in one go.

I keep my prompts scoped to drafting / refining specific components. If I feel stuck, I’ll use the chat as a rubber duck to bounce ideas off of and rebuild momentum. Asking the model to follow a specific pattern you’ve already established helps with consistency.

Recently been impressed with Gemini 2.5, it’s able to provide more substantive responses when I ask for criticism whereas the other models will act more like sycophants.

I want to believe that even though the context for that 2016 video was much different, he would still feel similarly about today’s AI outputs.

I don’t find joy in replacing artists wholesale. I think there’s something particularly soul draining about seeing so much online content becoming AI slop.

this is a bit unfair — a degree is required in almost any professional context.

I don’t have that piece of paper and while I’ve done OK in tech, there are structural barriers to advancement without a degree.

the majority are simply following what the system dictates. deviating from that system is difficult and not for the faint of heart.

genuinely curious what kind of demographic is leveraging Oracle for cloud products — all I’ve heard about them suggests long-term pain.

this incident certainly doesn’t help inspire confidence in their offerings.

the flip side to that “high” that comes with working super quickly for me has been a crash associated with a sinking feeling that I’ve outsourced too much.

Lots of discussion recently around how Next.js is “hard to deploy” without Vercel and don’t quite understand.

I’ve been using Next deployed to a container for years without issue — what are the major challenges here?

two things can be true:

- tech work is relatively privileged

- we, as tech workers, are losing out immensely due to a lack of collective bargaining

I look at the recent SAG-AFTRA strike as a reference: these are jobs with similarly privileged work conditions. However, there being a union meant they could inflict significant pain on industry players until their demands around AI training were met.

Meanwhile, we’re all just keeping our heads down hoping the market improves at the same time the biggest players are using our inputs to train even better models.

Just because our work conditions are better than other professions doesn’t negate the notion that we are being actively exploited to our own detriment.

Front-end engineering is incredibly driven by the “fashion” of the moment. The only framework that seems to have maybe overcome fashion is React.

Though React is plagued with its own churn around trends. Perhaps this is aging me a bit but I still can’t get my head around why we threw away class-based components with clear lifecycle methods.

I use hooks regularly now but it just feels so much less elegant than the perfectly fine solution that came before it.

there is, gofmt :)

If only we could set up subsidies for this infrastructure in the same way we subsidize ag and oil…

Thanks for sharing man. I think quite a few of us are fed up with the status quo in tech.

I was also at Microsoft until fairly recently, and although I didn’t feel like my immediate work was “unethical”, I’ve felt for quite some time that leadership is completely out of touch with workers. The copilot push in 2022, coinciding with firing the AI ethics team, is a prime example of actions that felt reckless.

I wonder what it’s going to take for tech workers like us to collectively say “enough is enough” to the grotesque avarice we see from tech leadership. I’m not holding my breath, and have essentially no desire to find another role in the industry.

Is this patent the only thing that is holding them back?

Or are there still quite a few challenges ahead and this is merely one roadblock removed.

so if someone makes a change to the system… there’s a person somewhere holding themselves accountable for the faults of the system, no?

really sad that this is what it's coming to... why are we having to play these games to run a simple search?

Perhaps being a PM for several years has helped, I’ve had great success speeding up my programming workflows by prompting Claude with very specific, well defined tasks.

Like many others are saying, you need to be in the drivers seat and in control. The LLM is not going to fully complete your objectives for you, but it will speed you up when provided with enough context, especially on mundane boilerplate tasks.

I think the key to LLMs being useful is knowing how to prompt with enough context to get a useful output, and knowing what context is not important so the output doesn’t lead you in the wrong direction.

I cannot believe “the US DOGE service” is a combination of words I have to read.

A meme coin representing the entirety of our government’s digital presence? Feels very unserious.

I really want to get off LinkedIn for this same reason. The happy-speak is so nauseating. The only reason I stay on is so recruiters can find me, but would be great to hear more of your experience in the job market without it.