1. Start with the ‘brainstorm’ session where you explain your feature or the task that you're trying to complete.
2. Allow it to write up a design doc, then an implementation plan - both saved to disk - by asking you multiple clarifying questions. Feel free to use voice transcription for this because it is probably as good as typing, if not better.
3. Open up a new Claude window and then use a git worktree with the Execute Plan command. This will essentially build out in multiple steps, committing after about three tasks. What I like to do is to have it review its work after three tasks as well so that you get easier code review and have a little bit more confidence that it's doing what you want it to do.
Overall, this hasn't really failed me yet and I've been using it now for two weeks and I've used about, I don't know, somewhere in the range of 10 million tokens this week alone.
You will simply need a lot of GPU cores/VRAM. On my $4,000 Mac Studio M2 Ultra with 64GB of RAM, I can comfortably run deepseek-r1:32b, but a) load times can be annoying (i.e. if you are switching models for different tasks, or let them idle out) and b) you can certainly tell that it requires tuning of the context length, temperature, etc. based on what you need to do.
Compare that with the commercial models where a lot of that is done on a large scale for you.
Yeah that makes sense. Once the model is loaded though, does it work well in comparison to the commercial models? Do you find that the local models hallucinate more, or don't give the same response quality?
Willing to relocate: Not for at least 1-2 years
Technologies: Kubernetes (cloud & bare metal), Terraform, AWS, Go, Python, Ruby on Rails, React/Next.js, MySQL/PostgreSQL, Redis, nginx & Apache, various cloud/video streaming/realtime tech, and more
12+ years as a platform engineer -> SRE at a number of top-tier startups or PaaS companies. Currently seeking my next senior or staff level SRE/DevOps/Platform Engineer position from Chicago or beyond!
A few projects I'm proudest of recently:
* Successfully planned & executed a six-month migration to EKS most recently at Sprout Social using custom-built pipelines which improved overall reliability and scalability for 150-ish microservices.
* Achieved a 75% reduction in AWS costs for the Moovweb (now Edgio/Akamai) Mobile Edge CDN by optimizing Node.js runtimes for use with Kubernetes.
* Helped develop two internal Go web apps used by the AWS EC2 CSET team to assist with enterprise client support engagements and dependencies.
* A couple years ago, I was instrumental in iteratively upgrading Hotel Tonight's core API backend from Ruby 1.8 to Ruby 2.3 with zero downtime, reducing endpoint response times by 45% and improving overall system performance.
Looking forward to hearing about any opportunities! Thanks!
You can rent a Tesla from Hertz for $40-50 per day, and they are currently including the Supercharger fees (at least I didn't get charged for it the last two times I took one).
Molly was incredibly kind to me. I met her a few times at conferences, and she always seemed to be bright, quirky and impossibly energetic. I am sorry to see this happen to such a lovely person. I hope she rests in power!
If you want to run them in an environment which reduces payload size (i.e. AWS Lambda) you really do, I'm afraid. This is possibly great news for Go users on AWS :)
The gay community has flawlessly foreshadowed the culture of the general populace for hundreds of years. People ignore this for stupid ideological/emotional reasons. If you want to see what the future Tinder looks like, look at Grindr today.
I really really don't think that's going to work...
Women have enough issues being harassed on the dating apps as it is...
Grindr conversations are very... forward (and would be seen as harassment by most Woman I have chat with about Grindr)
In reality the only reason the Grindr way "works" is because there is no societal power difference between everyone on it vs a hetero focused dating app.
It has always amazed me how well Grindr works, and yet every dating app catering to lesbians is a complete trainwreck. No hetero power dynamics on Her either but everyone on there seems more than a bit unhinged (as compared to the crowd on Tinder/Bumble) and the entire user experience is terrible. But on apps that allow hetero dating there are sooooo many men posing as lesbians by changing their gender settings or "bicurious" women who aren't actually looking to date women but changed their settings to show their profile to women just for fun because they like swiping and there's such a low barrier to entry. In case anyone reading this works there: I know I'm not the only one who really wishes Grindr would create a women-only clone of their app!
The same type of technical app may still turn out quite differently if it's for a vastly different demographic; it may be worth comparing outcomes to some relevant baseline metrics, such as the female divorce rate being >2x that of the male divorce rate in homosexual marriages in most countries https://en.wikipedia.org/wiki/Divorce_of_same-sex_couples
Very true. Based on personal experience I'm convinced that the female/female divorce rate is so much higher not because women have worse relationships with other women but because women have a tendency to rush into "serious" relationships that ultimately don't last. This would certainly lead to a different dynamic than you see on Grindr. The lesbian dating style is very much:
1. meet girl you like
2. almost immediately become exclusive
3. date "seriously" for several months, effectively skipping the early phase of the relationship where straight couples tend to play mind games or continue casually dating other people
4. break up
5. repeat
My gay male friends tend to casually sleep around a TON until eventually settling into a very serious long-term relationship, while straight couples are somewhere in the middle and have a much longer casual dating phase before making relationships official (at which point the relationship is more likely to last). When you take men out of the dating equation, the vast majority of women become serial daters who hop from one semi-serious semi-long-term relationship to another. It makes sense that a number of those couples would get married and then quickly get divorced before an equivalent straight or gay male couple would even get married in the first place.
It's actually quite interesting to observe these dynamics IRL; makes me feel a bit like a dating anthropologist.
I just don't see societies impressions about gender roles changing anytime soon.
Woman are still taught that they have to be hard to get and can't come off as "easy".
And Men are largely taught that getting a Woman is a conquest or somehow contributing to them "being a man".
Also I do feel the need to point out that on Grindr much of the same things that would be considered harassment on other apps is still there, it is just assumed to be part of the Grindr experience and not seen as harassment (by most at least).
(Purposefully not throwing in Trans here since that further complicates what I am trying to say and makes the assumed power differences even more problematic)
Also btw, if it isn't clear by now. I am saying this as someone on Grindr and other gay apps.
My profile isn't necessarily shy, but all of these things have happened and yes, it is expected to be part of the territory.
1. People calling me by my name in the first message. Because they know me and think its funny, despite them having no profile picture nor identifying themselves, etc.
2. Nudes. Lots. Unsolicited. Every angle, every sight, no crevice left un-illuminated.
3. A level of straight-forwardness that I'm not even comfortable fully describing here, as an example, on a throwaway. Near-interrogations about interest in fetishes, positions or play-by-plays of what is going to go down on a given meetup.
4. Calculation. Every gay man on Grindr knows the game. "You interested?". If the answer is "yes", then the answer is "maybe" - they're talking to someone hotter and seeing if it will work out. If it was going to work out, you'd have swapped numbers or one person would be on the move. There's games here too.
5. Bluntless. Such a time saver when someone says "not interested, good luck hunting". Rudeness. Some people have less tact when doing it.
6. Invasive questions. Are you clean (a bad way of asking if STD free), are you clean (a somewhat fair question about your ability to receptively bottom), how big is your penis, how fat are you, etc. Though I have never been asked how tall I was...
Anyway, I know lots of men that are completely turned off by this and I can fit more fingers in my nose than it takes to count the number of women (whom I know) that would opt into this experience. And yes, these aspects are integral to the Grindr experience, it wouldn't be Grindr and it wouldn't be popular without it.
Grindr is cool, but I feel like Tinder has already lowered the barrier for casual sex for straight people.... and uh, that's not solving any of the long-tail issues being discussed in this thread.
As someone who accidentally met their partner of 6yrs while looking for casual sex on OkCupid, and someone who is and was just on Grindr earlier... no. No.
Grindr's great. I'm proud of us queer folk for eschewing the shame and ironic "pomp and circumstance" of sex culture in the US, but, in my opinion, Grindr is NOT an example of how to fix dating apps for the masses, or for straight people.
(Though, and I don't say this to be mean or directly to who I'm replying to... I can certainly see why a straight man looking for sex would envy Grindr. At the same time, talk to any queer therapist or psychiatrist in the [large city] metro area and ask them about the mental effects of Grindr/Scruff on the LGBT community. It's a double-edged sword. What does it mean when there are literally thousands of horny men around you and NONE of them are talking to you?)
That all having been said, while reading this thread and seeing the number of men saying "I get dates and I'm not attractive", I'm realizing that either (1) the bar really just is lower for gay men on Grindr or (2) I'm better looking than I give myself credit for...
Also, I can't +1 enough what `nerdjon` said. Really condenses it down nicely.
Which makes perfect sense, because for woman reproduction/sex is a high-investment affair, whereas for men it's not. You see this all across the animal kingdom where males do all sorts of stuff to poach females.
Humans are really no different, and you can't just override billions of years of evolution with some anti-conception and an app.
> you can't just override billions of years of evolution with some anti-conception and an app
Why not?
The influence of evolution here is only relevant insofar as it affects decision making i.e. brain chemistry. It's reasonable to assume that the combination of (1) cultural memes, (2) evolutionarily novel experiences created by tech, (3) visual stimulation on screens, (4) learning there are no consequences, and (5) a literal drug that you take every day, also has an influence on brain chemistry. Maybe to such an extent that it overcomes some inborn tendencies.
There's no practical difference between a brain state caused by a gene versus by something else, for example, a drug. People have all sorts of brain states that betray evolutionary advantageous behavior--just look at a mental hospital. Or what about people from different tribes all taking the bus together without breaking out into warfare.
The fact that some behavior has its origins in evolution does not make it impossible to manipulate.
The article did say that it as a straight male he was not going to talk about LTBTQ+ apps.
That being said, Grindr by its very nature removes the balance issue in this particular sense. But Grindr is also a "dating" app by App Store description alone and is really not its primary use.
Have I had dates off of Grindr? Yes... But they are the exception. So looking at Grindr as a fix is not the way to go, it was very clearly not designed with that in mind (If when I talk to any of my straight male friends about Grindr is any indication)
That and... Grindr has its own major issues fueled by how its designed to be very much not a dating app.
Grindr has been moving more and more into the "dating app" territory. As of last month they introduced the ability to pin multiple Spotify songs to your profile just like on Tinder. That doesn't aid hooking up, that aids self expression and find other users you relate to.
Yeah they have added some features that are more tied to friends or dating.
But those are by far the exception after of years calling themselves a dating app.
It wasn't that long ago that they added the "Accepts NSFW" profile option. Expiring photos (if you do ultimate).
The Spotify one actually took me off guard when I saw that added, struggled with seeing the point until I remembered that some people do actually use it for another purpose.
My biggest issue with calling Grindr a dating app though is its focus on superficial (great for hooking up). But your profiles have a severe limitation on how much information you can put on them. Compared to something like OKCupid that makes me think of old myspace pages which could have a near unending amount of information.
An update or 2 ago they finally added tags which has been a bit helpful, but still only good for quick things.
I still haven't figured out what tags are useful for tbh. Are we able to filter by tags? Are there any common sets of tags besides the three they suggest (e.g. #sayhi)?
Sadly you can't filter by it, but it does kinda make sense since it is freeform so not sure how valuable search would be for that.
I just have #gaymer and #geek for mine
1. Start with the ‘brainstorm’ session where you explain your feature or the task that you're trying to complete. 2. Allow it to write up a design doc, then an implementation plan - both saved to disk - by asking you multiple clarifying questions. Feel free to use voice transcription for this because it is probably as good as typing, if not better. 3. Open up a new Claude window and then use a git worktree with the Execute Plan command. This will essentially build out in multiple steps, committing after about three tasks. What I like to do is to have it review its work after three tasks as well so that you get easier code review and have a little bit more confidence that it's doing what you want it to do.
Overall, this hasn't really failed me yet and I've been using it now for two weeks and I've used about, I don't know, somewhere in the range of 10 million tokens this week alone.