kamranjon

Yea this is some real tech bro up ranting at 3am kind of vibes - definitely AI assisted copy too…

Kind of a bummer cause I got excited when it started out with the promise of a new browser and then fizzled out with another “we shoved AI into something that absolutely didn’t need it” types of sales pitches.

Just because you added AI to something doesn’t make it valuable, and in most cases I’m going to assume you’re lazy and what you built is bad.

I’m somewhat confused by the distributed story for tiger beetle. In the docs here: https://docs.tigerbeetle.com/concepts/performance/#single-th...

It says: “TigerBeetle uses a single core by design and uses a single leader node to process events. Adding more nodes can therefore increase reliability, but not throughput.”

How does this work with multi-region? Does this mean that regardless of where in the world your users might live, they need to make a request to the leader node for a write to occur?

Is distribution purely for redundancy?

Right but libsql(Turso) uses HNSW - so i'd be curious to know how performance of sqlite-vector compares - they do say they "Use a brute-force-like approach, but highly optimized." - which to me, would be very interesting to see compared with a HNSW approach.

I'd be interested to understand the query performance when compared to the HNSW implementation (Turso?) they mentioned. In general search performance is more important to me, and I don't mind having an increase in insert overhead to have very fast vector search.

If you think that inference at OpenAI is nearly free, then I got a bridge to sell you. Seriously though this is not speculation, if you look at the recent interview with Altman he pretty explicitly states that they underestimated that inference costs would dwarf training costs - and he also stated that the one thing that could bring this house of cards down is if users decide they don’t actually want to pay for these services, and so far, they certainly have not covered costs.

I admit that I didn’t understand the Pro plan feature (I mostly use the API and assumed a similar model) but I think if you assume that this feature will remain free or that its costs won’t be incurred elsewhere, you’re likely ignoring the massive buildouts of data centers to support inference that is happening across the US right now.

Can this be interpreted as anything other than a scheme to charge you for hidden token fees? It sounds like they're asking users to just hand over a blank check to OpenAI to let it use as many tokens as it sees fit?

"ChatGPT can now do asynchronous research on your behalf. Each night, it synthesizes information from your memory, chat history, and direct feedback to learn what’s most relevant to you, then delivers personalized, focused updates the next day."

In what world is this not a huge cry for help from OpenAI? It sounds like they haven't found a monetization strategy that actually covers their costs and now they're just basically asking for the keys to your bank account.

Is Tesla robotaxi actually even close to Waymo? I thought they still needed someone in the car with the robotaxi and Waymo had been operating fully autonomous for quite a while already. My understanding is that this is one of the reasons people really like Waymo, it’s like a private ride.

Gotta love the employees feet pics folder found in the trash bin: https://posthog.com/feet-pics

This is great - thanks for posting - I can definitely use Pion in the game I'm working on and it's pretty great that it can compile to wasm and work on the frontend (for the clients)

I sit on a few indie hacker rooms with folks that slurped up the vibe coding koolaid immediately and they've been posting about their 100x output for over a year now - but none of them have shipped anything remotely successful - so I think at least anecdotally I somewhat agree.

I also just overall haven't seen a huge influx of new and useful apps and websites as you'd expect based on how tech leaders are talking about the virtues of these new tools.

I don't think AI has really found it's niche yet, and I think it's going to be much subtler than most people think, it's going to be the tools that integrate features like translations and summarization and speech to text in ways that are seamless that end up sticking - all of this other noise is just marketing hype.

Don’t all of the M series chips contain neural cores?

That statement could mean many different things

It’s good - I’m working through it right now

How did it compare with Gemma 3 models? I’ve been impressed with Gemma 27b - but I try out local models frequently and I’m excited to boot up your 70b model on my 128gb MacBook Pro when I get home!

Hmmmm… what is your opinion on the examples showcased here vs the ones on the Dia demo page?

https://yummy-fir-7a4.notion.site/dia

I am not sure why but I find the pacing of the parakeet based models (like Dia) to be much more realistic.

I read the whole thing but I got a strong AI vibe about half way through, felt like it was significantly padded with extra words to extend its length. Kind of annoying that blurb wasn’t placed at the beginning of the article…

Does anyone know why this says "No javascript, no stylesheets, no trackers. Just your words."

But when you look at the github repo here: https://github.com/HermanMartinus/bearblog

It says the repo is 40% CSS and 31% Javascript? Am I misinterpreting what this means in this context?

Often in go I’ll create some data structure like a map to hold the new value keyed by the original index (basically a for loop with goroutines inside that close over the index value) - then I just reorder them after waiting for all of them to complete.

Is this basically what Java is doing?

I think that maybe the techniques in this article are a little more complex, allowing you to optimize further (basically continue working as soon as possible instead of just waiting for everything to complete and reordering after the fact) but I’d be curious to know if I’ve missed something.

Has anyone used the Todoist app? What are the apps that you find the most helpful in supporting ADHD - I am thinking to get this app, but also have never used Obsidian before, would love to hear what others are using.

Edit: Oh also want to mention that I generally prefer privacy oriented apps - so if there is something, even paid, that will keep my notes on device without sharing them with a server I'd love to hear about it.

Just a heads up, Ive taken stimulants on and off as a treatment for ADHD for many years but my body/emotional health always felt compromised as a result. I've recently started on a non-stimulant ADHD medication called Atomoxetine and so far it has not had the emotional blunting, irritable effect of stimulants at all, and I haven't noticed any negative effects so far. It seems to help me get over the hump of being able to start things and stay with them which has always been my biggest downfall. We will see if I stick with it, but just wanted to mention that there are alternatives. There is also another medication called Guanfacine that I may try if this current medication does not work out - I don't think I can go back to stimulants.

What are the implications of this for GrapheneOS? Because it’s based on android, will that project die off?

How would this work well? I kinda thought that the whole benefit of torrents was being able to parallelize the downloads in chunks across the entirety of the file - wouldn't this negate that?

A really good book I read recently was Game Programming Patterns by Robert Nystrom - I think it has helped me, as a more traditional software engineer, to dip my toes into game development in a way that doesn't feel like a spaghetti plate full of hacks and tricks. I recommend it.

Awesome way to demonstrate exactly what your library does - also the library itself is a great concept - was just looking for something just like this the other day.

I would hope that there would be some possibility for folks like this to utilize platforms like Patreon or GitHub sponsorships to help pay their bills - they are obviously motivated and productive coders and also talented writers.

In my experience there is sort of a discrepancy often times. Someone discloses a severe mental health diagnosis and the company just goes on pretending they are just like every other employee - making no changes to accommodate them.

On the flip side, employees with severe mental health disorders often want to be treated exactly the same as other employees - so it seems a bit of an un-winnable situation.

Would you consider writing a blog post about this experience? I'm incredibly interested in learning more details about how this unfolded.

wow that timeline is a bit of a roller coaster:

"In 2022, my project HowToCook became enormously popular on GitHub, causing me to spend significant time maintaining it.

Unfortunately, my relationship with Lily changed. Her arrogance and disrespect led to us breaking up. I learned that people change."

...

"After moving my account and files to Suzhou, I explored GPU computing, built a Boeing 737 simulator in Suzhou Center, and reignited my relationship with Lily, looking forward to our future together."

I get the feeling that it is not actually the tech involved in the printers that distinguishes Bambu from Prusa, I think it's more about the supply chain and the distribution network. If I go to Prusa right now to order a core one printer from the US it tells me this: Estimated lead time 1–2 weeks

That means it's not even going to ship, from Europe, until then... And guess what? The shipping can range anywhere from 60$ to 300$ depending on the printer... Bambu has warehouses on US soil where they maintain stock of frequently purchased items and their printers/parts can be at my door in a matter of days with shipping ranging from 20$ to 100$ for their largest printer. It seems small but when you run a business that is reliant on 3d printers - these things matter. I think Prusa just honestly needs to focus on their distribution chain.

Like I really have considered Prusa printers for my business many times, but they either have had crazy lead times/shipping times or the prices out the door just don't make sense.

You can get consistent voice by providing a sample - and yea the timing stuff is what you have to work around - have to basically chunk your inputs.

The best open one I've found so far is Dia - https://github.com/nari-labs/dia - it has some limitations, but i think it's really impressive and I can run it on my laptop.