As someone who works on systems at this level, believe me, it’s a learnable skill. And at least an intellectually valuable one I think too. Even if you never really need the knowledge for the things you do, there’s a nice feeling that comes from seeing something done at a high level and understanding how that makes its way down into the system and why those design choices were made.
If I were more money motivated I’d probably be building CRUD apps too. I just like weird puzzles XD.
I’m always kind of envious of the people who were able to lose weight on GLP-1 drugs. I lost a bunch of weight a few years ago, and still need to lose a lot more (430 lb -> 330, goal 240), but I fell out of the good habits for, well, no good reasons…
Decided to try Ozempic and was on it for about 6 months. Didn’t do a single thing for my appetite unfortunately, even on the max dose.
Sample size of one here, but if you’ve got mental health struggles that feed into your eating patterns, GLP-1s might not help with your weight problems.
Wegovy/Ozempic didn’t do anything for me for months. Then my doc put me on Tirzepatide+Phentermine combo and I forgot what being hungry even feels like.
I couldn't hit my macros on tirzepatide - couldn't get enough protein without feeling sick to my stomach. One of the reasons I swapped to reta.
(though as a general note for anyone reading, just getting enough protein isn't enough - you need muscle stimulus too. Getting enough protein will help reduce the amount lost but if you really want to stop it, you gotta do resistance training)
I went down from 390lbs to 240lbs gradually over 5 years. I have maintained a weight of 240lbs since then (6'1" tall).
The first year was the most dramatic loss of 100lbs. I was miserable and didn't know what I was doing other than counting calories. The rest of it was more considerate of total nutrition, and that's what made my good eating habits stick.
I say this because while I'm not a doctor I think GLP-1 is probably unnecessary for the vast majority of patients. Better food and information is more available than ever before.
I would strongly advise to watch your A1C and get out of the diabetes danger zone if you are. Most people can drop a few percent in as little as 6 months and it makes a massive difference in mental health. Blood glucose has a direct impact on the brain and overall cardiovascular health. If you drink alcohol, you might want to take a break also to let your liver/kidneys/pancreas do their jobs properly and restore insulin sensitivity and other hormones. Look into the "fruit paradox", and more generally get a good salad in for lunch to address nutrient deficiencies. Not crappy salads either. You're not a rabbit. Treat them like the amazing sandwiches without bread that they are.
Sounds like old advice, because it is, but I find people aren't listening because they want to more deeply understand why to do it and what the effects are. Convenience and unintuitive pricing are false bargains that get in the way of healthier habits. Focus on nutrition and not quantity. Change your groceries, change your life.
> I say this because while I'm not a doctor I think GLP-1 is probably unnecessary for the vast majority of patients.
We have mountains of evidence that willpower fails for something like 99% of everyone, which is far from a vast majority. I applaud anyone's efforts to become healthier, however (though 240 at 6'1" is still obese, I would still explore medicine if I could not get any lower "naturally").
Thanks for the reply. Your perspective framing this as "willpower" is precisely what I'm concerned about.
I didn't need any willpower to do this and I'm not even humblebragging nor think of myself as a tough guy. I'm saying that healthy habits are simply a matter of understanding. If someone wants to take GLP-1 on top of that, it's their call. Many seem to be under the impression it's so vital for their specific situation to lose weight or avoid a heart attack and I think that's plainly false. We shouldn't be feeding fear, and humans aren't that unique.
I did not change my diet. If anything I just added more variety with a specific intent and it worked. Even just changing the order in which one eats things (fiber before sugary foods) can make a big difference. Once I got the blood glucose under control all the strong cravings and eating mistakes basically went away on their own without my conscious effort. The body is all connected and driven by hormones.
Sorry, you're right. I meant that I did not make significant changes to my diet. My point was I didn't really change what I eat, but how I eat. I still hate certain vegetables like carrots, kale, brussel sprouts, etc. and just added more of the nutritionally equivalent and culinarily far superior vegetables I was already eating.
That's not willpower. That's looking things up in the USDA database and tweaking my existing recipes. Why force nasty carrots onto the plate when I can eat spinach, cantaloupe, pumpkin, sweet potato, etc.?
I guess I also didn't emphasize enough that I took things super slowly? Taking 5 years to do what I did is a really modest goal. I just wanted to manage risk with minimal change. This is the pareto principle in action.
If we're really going to argue over stats, the effects of GLP-1 is meaningless noise in comparison and probably way harder to commit to. I just wanted to eat good and not feel like shit all the time. Isn't that what everyone wants? What if instead of there being "one weird trick" or a "miracle drug", we consider that basic nutrition is simply misunderstood and full of hundreds of weird tricks that are proportionally much easier to implement and they're damn tasty too?
I'm not knocking anyone meeting their goals without GLP-1s. It's obviously possible in absolute terms - people have been making great body transformations for as long as we've had fat people.
But everything you did, plenty of people try to do and fail at it. You are making it sound like this is all it takes and that it's easy. It might have been for you! But it might not be for other people.
The fact of the matter is the overwhelming majority of people that are obese and go on GLP-1s have tried other interventions before and failed at them. ~70% of all obese people have tried to lose weight in general, ~50% have recurring attempts, and while I don't have stats to back it up I am confident that the sort of people who are willing to go and inject themselves every week are the sort of people that have tried to lose weight in other ways.
> probably way harder to commit to.
A subcutaneous injection once a week is nothing. Dealing with constant food noise? I could maintain that if the rest of my life was stress free, and that's how I would drop 30lb. Once stress came back? So did the weight. Because for me, rearranging food doesn't matter if I still can't stop thinking about it even if I'm not actually hungry.
I'm on reta. It does barely anything to suppress my appetite - physical hunger has never been my issue. And I can easily eat however much I want - most days I am below 2k calories, but Saturday was an annual event with friends and I'm sure between food and alcohol I was probably at 5k calories for the day. But what reta does, is absolutely murders my food noise. I don't think about food constantly. I don't go eat because I got bored. The only thing I have to commit to for it is, once a week, put a needle on my injector pen, twist the dial to the right dosage, poke it into a spot where I still have subcutaneous fat, depress the twist top. Once a month I reconstitute a new vial.
I do at least one a day, sometimes up to four if things happen to line up exactly right.
Even four subq injections amounting to around 2ml of stuff is nothing, doing all four of them after a shower takes about as long as brushing my teeth.
If you use correct technique and good quality needles, you will feel essentially nothing. If your needles are not sharp enough, there might be very slight discomfort when initially piercing the skin.
My doctor, who is on the older side, told me that he went through his records when GLP-1s started being prescribed for weight loss. He wanted to calculate what percentage of his patients (a) he had advised to lose weight, (b) reduced their weight to healthy levels, (c) and kept it off.
From the starting population of overweight people, only 3% of people dropped down to, and stayed, a healthy weight.
that's too bad. I'm microdosing tirz -- 1/2 the "starter" dose they use to titrate you on, so essentially 1/4 of the therapeutic dose. Totally helped me reduce appetite driven snacking and eating. I still enjoy food, but now I leave half a beer left on the table. No problem. Makes you wonder what free will is like.
I am in a totally different setup than you though - 225 weight, goal to get down to 180 (my Army weight was 170.) I exercise a lot, and lost 30 pounds through diet two years ago. But it was the hardest thing I've done since Ranger school! So paintful to not eat enough. Gained it all back. I'm in month one of tirz and it is as easy as pie.
Yeah, the 18 months it took me to lose that first 100 pounds was about the most miserable I've ever been in my entire life. Constantly hungry, constantly thinking about food, exercising 3 hours a day, having to sit out of social events because more than half of them involve going to restaurants, etc.
It really opened my eyes to just how food centric society is.
That's unfortunate! It might be worth checking out Tirzepatide or Retatrutide once it is released. The GIP and Glucagon receptors might be better targets for you, even if the GLP-1 receptor seems to not help.
It lets me more or less skip a meal but holy hell I am craving sugar more than ever. On the whole I'm cutting calories and have lost a lot of weight, I just wish I didn't want sugar this much.
I’ve always been more of a savory kind of person myself. I’d take biscuits and gravy or a steak over sweets any day!
But I feel you on sugar. Took me a long time to cut sugar cravings. A decade ago I cut regular soda out of my diet, which a few years later led to me cutting out pretty much anything sweetened. Realistically it wasn’t the sweetness for me, it was the “mouthfeel” or doing something with your mouth. Just straight sparkling water satisfied the entire craving for me.
The hardest thing for me to give up / heavily cut back on was fried things. Maybe that’s the result of my parents using french fries as the reward food when I was a kid…
Try allulose-sweetened stuff. Allulose is a sugar your body doesn't metabolize like sucrose. It has zero calories and does not increase your blood sugar. It's a component of maple syrup and so does taste a bit maple-y, but better than most artificial sweeteners and even stevia leaf extract (stevia and aspartame have a "tang" to them I dislike).
I really had thought (with no research) the correlation between mental health and glp1 effectiveness went the other way around. Thank you for this check-your-biases moment, you probably just saved me a ton of embarrassment down the line, if these drugs ever enter my life.
I don’t think there is remotely enough data on the subject to make any confident statements either way yet.
I think the only very confident thing I can say after watching and helping dozens of folks get started on these drugs is that everyone’s biology is vastly different.
I have friends who have lost close to a hundred pounds on the starting doses of their chosen GLP-1. I have other friends who barely lost anything after a year at max dose. Some of these people in both groups are highly motivated to lose weight and some are simply taking the drug as a magic fix and expending zero other effort into changing their lives. Some have very difficult mental issues and relationships with food, some have very few hangups on the subject.
I have never been able to predict with high confidence how any particular person is going to react to taking them. By and large the results are close to magical for the majority of folks, and there may be some correlation with folks who combine the drug with other lifestyle changes - but those are just general averages I see and certainly not scientific.
You missed out on both of the weight suppression tricks, which really does suck. Appetite suppression (or reduction of food noise) is pretty useful, but GLP1s also tend to punish you mightily if you overeat. For me, even if I were hungry, overeating will make me hurt for hours. I could not gain weight on this even if I wanted to.
There are some difference, too, between the various drugs. I never tried ozempic, I went directly to tirzepatide (zepbound). And then to retatrutide. I will say that reta is in some ways the most interesting, because it has less appetite suppressing activity than tirzepatide (this is common, not just me), but it still cuts my stomach capacity quite a lot, and ramps up my metabolism. I had stalled at about 90 pounds down with tirzepatide, and reta immediately knocked off another 15. I track calories, and I had changed nothing. Felt more hungry, still lost more weight. Wild.
From one rando to another, I recommend trying tirzepatide. Or try semaglutide again but stacked with cagrilintide -- some people get pretty great results with that, similar to tirz.
Yeah I might just have to go get it out of plan. Kaiser covers Ozempic, but none of the Tirzepatide based medications. (Edit: looks like that might have changed)
I am a big guy (6’4, 330 lbs), but I was amazed that Ozempic just seemed to do nothing. I was having the gastric side effects, but I could still eat 3000 calories a day if I cheated without feeling anything.
Similar experience here with Tirzepatide. Overeating is punished swiftly and painfully.
If it works for you, look into getting one of the 15mg pens and counting clicks in order to get more doses per vial. I've been on the one pen for 3 months now and it's still got plenty of juice left.
One of the quirks of buying brand name GLP1s in the US is that we don't get the dial-a-dose pens, every autopen is one-shot. Some people disassemble them to get multiple doses, but at that point you might as well get the cheaper brand name vials or go with compound or gray.
Lilly Direct sells zepbound in “single use” vials you make draws from. Very trivial to add bacteriostatic water to them and do some simple math to divide the dose. I have a few friends who do this.
You can also take apart the pens and do the same thing, but it’s a lot more involved and you’ll need to source some sterile reusable vials for it.
When I was on Ozempic in the US (Bay Area), it was a dial-a-dose Ozempic branded pen. Came with 4-6 single use needles you’d screw onto the end before use, and discard into a sharps bin after.
Which do you like? Barebells salty peanut and chocolate dough over here. Though the sugar alcohols certainly aren’t great for you either, I think they were recently linked to stroke risk
I'm a Barebells Coco Choco "fan", though I'm aware of the stories around sugar alcohols. I think those bars are way too sweet anyway. They could use far less sweeteners. Would love to hear about more responsible options.
Thanks for the link. I also hate the sugar alcohols.
However:
> But allulose isn’t approved for use in Canada or Europe. There, it’s considered a “novel food,” which means it hasn’t been available long enough for sufficient testing, according to those governments’ standards.
> And it’s important to know that the FDA’s GRAS status doesn’t mean that allulose has been rigorously tested.
> “We don’t have studies regarding the safety of allulose at this time,” Dr. Hazen shares. “But if it follows similar trends to what we see in some other sugar substitutes that are sugar alcohols like erythritol, I would suggest there’s reason to be cautious about how much of it you consume.”
I don't consume more than like a few grams per day to sweeten things like coffee, as well as the 5g or so in a single Munk bar. I seem to tolerate it well.
To add to what you said, it’s also nice to be able to keep it within one API that’s platform agnostic when possible.
Sure we’ve had the ability to keep the pipeline on GPU for awhile, but it usually required platform specific API bindings to convert to a platform specific descriptor (handles on windows, IOSurface on macOS, dmabuf on Linux), which you then had to pull into a platform specific decoder/encoder API (DXGI, WMF, AVFoundation, VAAPI, etc.), and then all of that again but in reverse to get the surface back into your 3D API.
This whole thing just makes life easier for everyone.
Regardless of how you feel about Wayland, its creation set off _massive_ improvements across the entire Linux graphics stack.
For those of us who were using Linux on the desktop in decades past, remember when you couldn't use a GPU without X running? Remember the days when you needed an X session running in order to use CUDA or OpenCL? Remember the days when the entire graphics driver lived inside of X? When display server issues caused kernel panics? Remember the days when you couldn't share a hardware graphics surface between processes? When it was impossible to get hardware acceleration to work offscreen?
Wayland's aggressive stance on "it doesn't work on platforms that don't fix all of that" is one of the only things that pushed the stability and flexibility of the graphics stack on Linux forward.
I don't really think anything less than saying "We the X developers are going to stop X development and X is going away" would have been enough to push graphics card vendors to actually rework the drivers.
People cursed the name for years, because it exposed all of the terrible, glitchy audio hardware drivers and refused on general principle to work around the issues to the degree that previous audio solutions had. And the result was that while the experience was inconsistent and buggy for years, it did eventually drag the Linux audio stack into a better place.
Comparing JACK and PulseAudio is like comparing apples and oranges. And honestly, JACK by itself is unsuitable as an audio server for general desktop usage.
JACK doesn’t support device hotplug (ya know, connecting and disconnecting a headset, something most of us do) and it also doesn’t support multiple applications generating audio without the user having to configure how audio is mixed.
JACK is designed for low latency in environments like Digital Audio Workstations (DAW) where you know 1) what audio hardware is present at all times, and 2) what applications are going to generate audio.
Many people who use/used JACK ran a PulseAudio bridge on top of it for every application that wasn’t the one or two applications that needed ultra low latency audio.
PulseAudio had some major warts, but JACK wasn’t some panacea that did everything better.
The ALSA drivers for all the creative labs cards worked perfectly well. I never had any issues at all under ALSA, or under OSS before that.
I've had tons of issues with audio bugs once pulseaudio was introduced. To this day the most common solution to any audio issue I see is `pkill -9 pulseaudio`. And it solves the problem about 99% of the time.
It also used a large multiple more memory than Alsa + basically any existing plausible combo of mixer software on top of alsa. While doing nothing. For no clear reason. And chewed processor cycles, while doing nothing. Back when 50MB was a meaningful amount of memory, and most machines were still single-core.
It was plainly really poorly-architected, just looking at its resource use patterns made this obvious in a heartbeat.
It also introduced fun new audio bugs and indeterminate latency. Which still haven't gone away entirely in 2026. To such an extent that any time I have an audio issue, I reflexively `pkill -9 pulseaudio` and about 99% of the time the problem just vanishes.
On the first machine where I had pulseaudio foisted on me - an 800mhz single core Duron - pulseaudio used literally 20% of my CPU time...
...At idle. When no audio was playing...
...To do software audio mixing which my creative labs audio hardware was capable of doing better and for free.
When I filed an issue with the pulseaudio people, saying "hey, you're wasting 20% of my CPU time at idle when no audio is playing because you're ignoring the fact that I have superior hardware that can do audio mixing for free", they closed the issue saying that pulseaudio wasn't meant to be used in situations where you have dedicated hardware for audio mixing.
The unfortunate reality of building these home-brew CPUs is that almost all of the "medium integration" ICs are long out of production - things like the 74181 ALU slice, carry lookahead adders, 16-way register files, etc.
Makes doing things larger than 8/16 bit computers very complicated and usually very slow :(
It's one of the test layouts that I put in this perfboard layout program that I ened up making because trying to figure out wiring on both sides at once in my head melted my brain.
The problem is not that gaming GPUs are in demand, it’s that selling silicon to AI center buildouts is so absurdly profitable right now they just aren’t making many gaming GPUs.
If you can only get so many mm^2 of dies from TSMC, might as well make 50x selling to AI providers.
I just wish there was a stronger source on this. I am inclined to agree you and the source you cited, but unfortunately
> [1] This story requires some reading between the lines - the exact text of the contract isn’t available - but something like it is suggested by the way both sides have been presenting the negotiations.
I deal with far too many people who won't believe me without 10 bullet-proof sources but get very angry with me if I won't take their word without a source :(
> "Two such use cases have never been included in our contracts with the Department of War..."
While I agree with Anthropic's position on this regardless, the original contract wording does matter in terms of making either the government look even more unreasonable or Anthropic look a little less reasonable.
The issue is a subtle ambiguity in Dario's statement: "...have never been included in our contracts" because it leaves two possibilities: 1. those two conditions were explicitly mentioned and disallowed in the contract, or 2. they weren't in the contract itself - and are disallowed by Anthropic's Terms of Service and complying with the ToS is a condition in the contract (which would be typical).
If that's the case, then it matters if the ToS disallowed those two uses at the time the original contract was signed, or if the ToS was revised since signing. Anthropic is still 100% in the right if the ToS disallowed these uses at the time of signing and the ToS was an explicit condition of the contract, since contracts often loop in the ToS as a condition while not precluding the ToS being updated.
However, if the ToS was updated after contract signing and Anthropic added or expanded the wording of those two provisions, then the DoD, IMHO, has a tiny shred of justification to complain and stop using Anthropic. Of course, going much further and banning the entire US government (and contractors) from using Anthropic for any use, including all the ones where these two provisions don't matter - is egregiously punitive and shitty.
While the contract wording itself may be subject to NDA, it would be helpful if Anthropic's statements could be a bit more precise. For example, if Dario had said "have always been disallowed in our contracts" this ambiguity wouldn't exist.
It does not matter. If Anthropic had been precise in this narrow way, there would have been some other nitpick to raise.
You're trying desperately to find a way that things can be at least a little normal, and I really do get it. It would be great if such a way existed. But it doesn't. I recommend you take a social media break like I'm about to, take the time you need to mourn the era of normal politics, and come back with a full understanding that the US government is not pursuing normal policy objectives with bad decisions. They hate you and they hate me for not being on their side, and their primary goal is to ensure that we're as miserable as they can make us.
I'm in a weird spot where I do agree with your assessment of the core claim. But putting that aside, in the world where the DoW's claim _is_ correct -- I think you don't have any choice other than to designate them a supply chain risk.
Disregarding who is right or wrong for a moment, if the DoW are right (which I'm not personally inclined to believe, but we're ignoring that for the moment) -- how else can they avoid secondhand Claude poisoning?
Supposing they really want to use their software for things disallowed by Claude's (now or future) ToS, it seems like designating it a supply chain risk is the only way they can ensure that their contractors don't include Claude (either indirectly as a wrapper or tertially through use of generated code etc)
> designating it a supply chain risk is the only way they can ensure that their contractors don't include Claude
I agree that if the DoW claim is correct (and I doubt it is), then, sure, the DoW dropping Anthropic and precluding the DoW's suppliers from using Anthropic for any DoW work would be expected. However, the "supply chain risk" designation they are deploying goes far beyond that to block Anthropic use by any supplier to any part of the entire U.S. government for anything.
For example, no one at Crayola can use Anthropic for anything because Crayola sells crayons to the Education Dept. The DoW already has much less draconian ways to restrict what their direct suppliers use to build things for military applications. But instead of addressing the actual risk in a normal measured way, they are choosing to use a nuke against a grenade-sized problem. This "supply chain risk" designation is rarely used and has never been used against a U.S. company. It's used against Chinese or Russian companies when in cases where there's credible risk of sabotage or espionage. That's why that particular designation always blocks all products from an entire company for any application by any part of the U.S. Government, contractors and suppliers (which is why it's never been used against a U.S. company).
One positive thing I will say about this administration is that they have really drawn into focus the difference between de jure and de facto law.
My hope is that this gets us some real concern for things that have been defended with de facto arguments (i.e. privacy) going forward.
edit: Anthropic argues that your Crayola analogy is fundamentally incorrect.
> Legally, a supply chain risk designation under 10 USC 3252 can only extend to the use of Claude as part of Department of War contracts—it cannot affect how contractors use Claude to serve other customers.
> Anthropic argues that your Crayola analogy is fundamentally incorrect.
Yes, I just saw Dario's latest post with that more detailed info. My understanding was informed by news reporting in a couple different outlets but those reports may have been conflating the "supply chain risk" designation (under 10 USC 3252) with the net effect of statements from the pentagon and white house which go substantially further.
Even if it's not in the legal scope of 10 USC 3252, the administration has made clear they intend to ban Anthropic from use across the federal government. AFAICT doing that is probably within the discretionary remit of the executive branch, even though I believe it's unprecedented - to your point about de jure and de facto law.
To me, if there's a silver lining to all this, it's making a strong case for restricting executive branch power.
Edit to add: Per the Wall Street Journal's lead story (updated in the last hour): "The General Services Administration, which oversees federal procurement, said it is removing Anthropic from its product offerings to government agencies... Even absent the supply-chain risk designation, broadening the clash to include all federal agencies takes the Anthropic fight to a much larger scale than its spat with the Pentagon."
How would this risk be mitigated by signing a contract? Seems like “supply chain poisoning as treason” is probably not going to stopped by a piece of paper. You either trust anthropic or you don’t but the deal has nothing to do with it.
Isn't the point that they aren't entering into a contract with them, they are just ensuring that none of their still trusted suppliers repackage Anthropic without their knowledge?
I’m not sure, but I think you’re right. I was thinking about the logical implications of the. If they are a supply chain risk without a contract, how does the existence of a contract suddenly make them not a risk? Especially if the DoD strong arms them into a deal.
Because the act that the SCR designation would “protect” against is treason, so I don’t think people would care too much whether there’s a contract.
Also, Trump's own words complaining about being forced to stick to Anthropic's terms of service:
> The Leftwing nut jobs at Anthropic have made a DISASTROUS MISTAKE trying to STRONG-ARM the Department of War, and force them to obey their Terms of Service instead of our Constitution.
In this case, do you really believe that we should trust an EA less than this administration? EA as bad people is a stereotype; corruption, fraud, and breaking the law is the standard MO for this administration.
(Or maybe it’s catchier to respond glibly with “never trust a child rapist and convicted felon.”)
In this case, the choice is between the two apples, so I’d pick the one less obviously rotten. Sadly that is the current administration that operates in pure lawlessness.
I think a big question mark here, is whether anything said on Anthropic's side if in the framing of "We have a thing going on that we are trying to communicate around where a canary notice if it existed would no longer be updated"
As someone who works in the datacenter hardware / system software space who does robotics as a hobby - the craziest thing to me about ROS is how tied it is to Linux and how Linux-illiterate half the people using it seem to be.
It was like pulling teeth trying to explain to someone that they couldn’t use the Linux GPIO subsystem from python to accurately measure sub-millisecond events…
Linux is a general purpose OS, it’s typically tuned for throughput not precise timing. Sure there’s PREMPT_RT but it only buys you so much.
Waymos require a highly mapped environment to function safely in. Not to take away from what Waymo has accomplished, but it's a far more bounded problem that what the "self driving" promise has been.
If I were more money motivated I’d probably be building CRUD apps too. I just like weird puzzles XD.
reply