Location: San Francisco Bay Area, CA
Remote: Yes (also open to hybrid and on-site)
Willing to relocate: No (open to Bay Area roles)
Technologies: Kotlin, Jetpack Compose, MVVM, Clean Architecture, Coroutines, Android
Résumé/CV: https://www.linkedin.com/in/meaydinli
Email: Please refer to CV
Android Engineer with 12+ years building consumer mobile apps at scale (fintech, logistics).
Led the final phase of a complete Android rewrite, owned CI/CD and release process, and built observability pipelines (Datadog + Crashlytics) driving ~99.9% crash-free sessions. Strong focus on performance, reliability, and maintainable architecture.
Experience mentoring engineers into senior roles, leading design reviews, and aligning mobile architecture with product and platform constraints.
Given that for a non quantized 700B monolithic model with let's say a 1M token context, you would need around 20TB of memory, I doubt your spark or M4 will get very far.
I'm not saying those machines can't be usefull or fun, but it's not in the range of the 'fantasy' thing you're responding to.
I regularly use Gemini CLI and Claude Code, and I'm convinced that Gemini's enormous context window isn't that helpful in many situations. I think the more you put into context, the more likely the model is to go off into on a tangent and you end up with "context rot" or get confused and start working on an older no longer relevant context. You definitely need to manage and clear your context window and the only time I would want such a large context window is when the source data is really that large.
Context quality and relevance is indeed a major factor. But large size is not the core issue, although in unmaintained or poor relevance context situations a smaller window is going to blissfully forget the bad, and the good, sooner.
I'm running an M4 Max as well and I found that using project goose works decently well with qwen3 coder loaded on LM Studio (Ollama doesn't do MLX yet unless you build it yourself I think) and configured as an openai model as the api is compatible. Goose adds a bunch of tools and plugins that make the model more effective.
No, it is not "Hacker News" to paint a large swath of people as "ignorant". The proper way would have been to publish a detailed, technical analysis and present your ideas along with your proof to the greater community and facilitate a discussion.
I want to live in this world, where we're all part of a larger collective endeavor to discover the truth, to be more informed, learn continuously, and evolve together. That's the old-school philosophy and mission of science for the betterment of humanity.
My guy, there are government officials that are sharing photos of the Orion constellation stating that it's a drone swarm just standing there over their property...
I am one of those weirdos that like to work on "un-sexy" things behind the scenes which hopefully makes the life of my colleagues a bit better with every attempt. I don't think it is necessary for small teams, and may even be considered a waste of time, but once a team is sufficiently big and spread out, I think a few people working in the background, keeping watch of things, cleaning up after people, proactively improving stuff, creating and enforcing rules and standards is very beneficial and necessary for a team's next growth spurt.
Groups like that use vague language like that on purpose so that when they are cornered and questioned, they can say non-committal, vague things like that and get away with not caring about nobody but themselves.
So yes, if they really mean protecting the community along with young artists, they have to explicitly and clearly state it to be believable.
