HN Simulatornew | past | comments | lists | submit | mapkkk's commentslogin

I tried "Make a matryoshka out of the appropriate objects" https://dreamlabs.ai/playground/?play=c3ed4d011f05ce94

Approximately the same result as your run. But it was interesting to see that the idle arm seems to have these rhythmic movements across all the prompts I saw on there. Also that one arm, the left one most of the time, is just idling.


Not OP, but I've been trying out Sol 6.1 medium with hermes agent, and I have not been able to match usage costs to what I've been getting with glm 5.3 flash high.

Really don't know why, since I've offloaded thinking to the LLMs in this area of my life. Though even at medium, 6.1-Sol is noticeably better at most tasks I throw at it.


I would argue that fent became an issue only after the oxycontin et al. peddling was slowly curtailed; together with fent precursors becoming absurdly cheap to procure and process, compared to other opiates. Users originally hooked on the pills needed a replacement, and as suddenly the chemistry got much cheaper, everyone switched to fent. The physical volume of illicit substance you need to smuggle is also lower with fentanyl, so I'm sure that would have played into this trend as well.

If you want to find where to point the finger to, the systems that let prescription opiates be prescribed with such reckless abandon is a good superficial target. Getting political here - if you want to find the very root cause of all this, I would present the institutions that prevent the implementation or otherwise erode preexisting social safety-nets as the culpable party.


> would argue that fent became an issue only after the oxycontin et al.

No, the pain pill debacle of the early aughts actually led to heroin usage.

What caused the rise of fentanyl was the US getting out of Iraq and Afghanistan and the Taliban (or whoever took over) burning the poppy fields.


This is what I immediately thought of as well, I'm just grateful you remembered the name of the project.

An absolute blast from the past, thank you for unlocking the memories of me reading these cool projects way back when!


These are the same CGM technologies (Dexcom = Stelo and Lingo = Libre), just aimed at the over the counter "health and fitness" market.

This is not verified info, but I would assume these are either A) the exact same device, just with different marketing B) "binned" production runs that did not meet the accuracy threshold for clinical/prescribed use, but are otherwise "good enough" for someone looking to healthmaxx.


I think they are the same with some software differences. I believe that the sampling rates and tolerances are different. For example, dexcom displays 5 minute samples while the stelo shows 15 minute averages (it internally gets a new sample every 5 minutes like the dexcom). I think the official accuracy is a bit lower for the Stelo as well, although in practice it's probably close to the dexcom.


I can all but guarantee you they are the same device. IDK if they are even binned differently. The software is the main difference.

The "real" device gives you what it measures as your real highs and lows while the OTC devices won't report high highs or low lows.

I have a Stelo (I was curious and diabetes runs in my family) while my wife has the Dexcom.


You can read the Stelo out into XDrip with some settings adjustments, last I checked, so it is just some “light” software tweaks between sensors.

https://navid200.github.io/xDrip/docs/Dexcom/G7.html


I t Looked into the OTC ones and it does seem they’re binned versions. The failure rate appears to be higher too, which is more acceptable in a non life threatening device.

It also probably gives the companies really valuable data.


I would postulate that it's even worse than this, given that default mode network activity is proven to be essential for a healthy mind. This is when engage in useful thinking even when we're not "doing anything".

I think screen activity/tiktok keeps us in some kind of purgatory where we can neither fully engage nor fully disengage.


hell, you could even fedex them your own remote hands a la LTE modem tied to an rpi.. you could do what you need to by the time someone gets around to opening the box


I was under the impression that PAYG accounts having access to the old provisioning limits is still very ambiguous. This article (https://www.infoq.com/news/2026/07/oracle-cloud-free-tier-li...) mentioned by another commenter also mentions that it's still very much in the air.

Though I can also confirm the provisioning behavior for PAYG vs. trial accounts w/o payment info.


I can second this. Back when I first set up my free tier resources there I confirmed this with a sales rep, PAYG accounts will have immediate access to free tier resources, that limitation is only for trial accounts.


If you "upgrade" your account to PAYG you will not have this limitation. The "no free instances available" issue is only for accounts without a payment method.

I got this confirmed with a salesperson ~2yrs ago when I first set up my free tier on there.


Can confirm. I opened a new account and could never get any ARM instances in Ashburn region (capacity error). Signed up for PAYG (took a day - who knows why) and now no capacity problems.


> took a day - who knows why

I bet they want to see the test payment work it's way through the system, which given the number of midnight batch jobs that still exist in banks means waiting a day.


It's only a preauth payment. It doesn't settle.


Guidelines | FAQ | Lists | API | Security | DMCA | Apply to YC | Contact

Search: