HNHacker News
TopNewBestAskShowJobs

cloudie78

353 karma · joined May 8, 2026

submissionscomments
cloudie78··on Semaglutide linked to lower predicted dementia risk
It’s in broad terms weight loss related. More specifically blood glucose and insulin regulation which can be out of whack even if you’re not diabetic on paper. This is why you’ll see people claiming they’re not diabetic but still saw improvement from Ozempic and friends. Their insulin and glucose regulation was spoke. in casual terms “fucked”.

GLP-1 receptor agonists such as semaglutide act partly through the brain to reduce hunger, increase satiety, and lower spontaneous energy intake. Sustained caloric reduction produces weight loss, including loss of visceral fat around the abdominal organs and ectopic fat stored in tissues such as the liver. These fat depots are metabolically active and contribute to insulin resistance through excess fatty-acid delivery, inflammatory signalling, and disruption of normal insulin action.

As they shrink, muscle and liver tissues become more responsive to insulin, so the pancreas no longer needs to secrete as much insulin to maintain normal glucose levels.

This reduces the compensatory hyperinsulinaemic state that can exist for years before diabetes develops and decreases the probability that impaired glucose regulation will progress toward overt type 2 diabetes.

Those changes propagate into the cardiovascular system.

Better insulin sensitivity and lower visceral and liver fat are commonly accompanied by lower triglycerides, improved blood pressure, reduced systemic inflammation, and better endothelial and vascular function, thereby reducing several mechanisms that contribute to atherosclerosis and cerebrovascular injury.

GLP-1 receptor agonists also have a direct metabolic action independent of weight loss: when glucose is elevated, they enhance pancreatic insulin secretion and suppress inappropriate glucagon secretion, improving glucose control without forcing insulin secretion when glucose is low.

The overall effect is therefore a combination of an immediate GLP-1-mediated improvement in appetite and glucose regulation followed by progressively larger downstream effects from weight and fat loss, which together reduce the metabolic and vascular abnormalities associated with cardiovascular disease and, over much longer periods, potentially with dementia risk.

cloudie78··on Single log line is 49KB+ (ext4) / 110KB+ (btrfs) of systemd-journald disk writes
Why not just have a SQLite file and call it a day?

Also, why mmaped file?

cloudie78··on DeepSeek up to 1000% price hike is live
I don’t think you could’ve picked a shittier title.

Don’t bother with the link, here’s a quick rundown:

DeepSeek is raising API prices for its V4 models and introducing peak/off-peak pricing. For V4-Pro, uncached input goes from $0.435 per million tokens to $0.66 off-peak or $1.32 at peak; output rises from $0.87 to $1.98 or $3.96. The four-digit percentage increase comes from cached input, which was priced at an unusually low $0.003625 per million tokens and is rising to as much as $0.044.

Even after the increase, DeepSeek remains cheap relative to major frontier models. Claude Sonnet 5 costs $2 per million input tokens and $10 per million output tokens; Claude Opus 5 costs $5/$25. OpenAI’s GPT-5.6 Terra is $2/$12, while GPT-5.6 Sol is $5/$30.

So at peak rates, DeepSeek V4-Pro still costs about two-thirds as much as Sonnet on input and roughly 40% as much on output. Off-peak, the gap is wider.

cloudie78··on My Rules for Using Spreadsheets
Honorable mention: https://www.theguardian.com/politics/2020/oct/05/how-excel-m...

That one time NHS had data loss because of XLSX limits

cloudie78··on Codex in ChatGPT desktop app for Linux is now in preview
It’s called LSP, any good editor supports it.

In fact, VSCode’s support for Lean is via LSP anyways.

cloudie78··on Codex in ChatGPT desktop app for Linux is now in preview
Trust and VSCode don’t go together in the same sentence lol
cloudie78··on Codex in ChatGPT desktop app for Linux is now in preview
Quite on the contrary, in fact.

Sectors involved in, in no particular order: government, financial, telecom, energy (gas and electric), manufacturing (large european conglomerates), healthcare, logistics

In fact I was the one usually helping IT Teams put the torture devices and policies you’re subjected to in place lol.

Banks would usually give us laptops with smart card readers and only allowed access to their VPN from a limited set of public IP addresses.

Govt was a bit more lax, but same principles applied, was really fun doing any kind of work over a laggy VNC.

My favourite one was probably where you’d have to hop from one Azure cloud VM, to Citrix to a bastion windows box, off of which you’d RDP to the actual work VM only to putty to the actual damn Linux box you cared about.

Energy sector was more lax, they let us set up a GitLab instance and a small CICD setup to which we’d push the code developed on our local boxes.

Oh and the ones I mentioned - they usually ask us what tools we wanted on our boxes as well.

cloudie78··on ChatGPT Desktop (Codex Desktop) for Linux
If your hypothetical client as security measure ONLY permits VSCode on a dedicated box you’re obligated to work on and have no control over - then I pray for you.

And for your hypothetical client.

Especially as judging by your description they despite this allow any VSCode plugin to be installed alongside it.

> and I am not making my life miserable to work with editors lacking the specific tools

And that’s the self-imposed prison I was talking about, there’s always alternative tools. In my mind a specific tool is something like Ghidra.

cloudie78··on Codex in ChatGPT desktop app for Linux is now in preview
You always have a choice to use/learn/make a different tool.

It’s a prison of your own making.

I’ve personally adopted a process where I structure my projects to ALWAYS keep credentials and sensitive information in a separate directory.

The working copy gets rsynced to and from a dedicated VM with whatever $HARNESS.

cloudie78··on ChatGPT Desktop (Codex Desktop) for Linux
If you want to be security conscious don’t give it access to your system directly. Ask it to guide you through the information gathering part of the problem solving process and use your brain and judgment if it actually needs the information/data it’s attempting to access.

If you can’t make that judgement call because of lack of knowledge/familiarity it’s okay - spawn a new fresh chat, get up to speed on the topic first and then resume.

That way you accomplish three things: you learn something new, you fix your problem and you don’t let a Trojan do whatever the hell it wants to your box which among other thing contains all the browser sessions and cookies :)

cloudie78··on Codex in ChatGPT desktop app for Linux is now in preview
Reminder to everyone, treat these as trojans. Run them isolated from the rest of your system.

Give it a full desktop in a VM if you want to, just not direct access to your system.

cloudie78··on Amazon backs power plant that may become top source of US climate pollution
If only we had the technology and ability to produce enormous quantities of energy non-stop 24/7 365, no matter day or night wind or no wind.
cloudie78··on Poland now 6th-largest EU economy, ahead of Switzerland and Belgium
GDP is a scam and a shit metric. Natural disasters increase GDP and so does pollution because of the money spent cleaning it up.

Poland also has 4x the population of Switzerland.

Normalise the GDP and then let’s start having a conversation. Look at the PPP.

cloudie78··on Danish high schoolers will have to verbally defend written assignments
Hi, former student with social phobia and terrified of public speaking.

Today: adult with no problems doing public speaking in front of hundreds.

What fixed me: more public speaking and not having the option of avoiding the task.

cloudie78··on When online commenters detect my art as AI
Internet commentators are dumbasses. Nothing new here.

Point and laugh at them, and move on. They don’t matter, they’re idiots.

Source: I’m a dumbass.

cloudie78··on The Tungsten Market Is Warning of an Upcoming War
> Tungsten prices are already quite easily explained by China restricting exports and reducing production

And why, pray tell, would China restrict the production and export of tungsten, along with many other critical dual-use metals and minerals?

cloudie78··on Apple is getting this wrong
Scam Altman at it again.

PR is not a moat buddy.

cloudie78··on The AI trade now runs on borrowed money, and the lenders are repricing it
> There's no spare money to extract from the airline industry but it's still very useful. The same could be true for AI in the long term.

Of course it could, let’s start with making the models open weight and entirely open source. Fully publicly owned and not shaped to maximise profits for the shareholders.

Oh wait, Scam Altman entered the chat and turned a non-profit lab into the next biggest IPO vehicle the world has ever seen.

OpenAI launched as a nonprofit research institution. Its announcement explicitly said it wanted to pursue AI “unconstrained by a need to generate financial return,” produce value for everyone rather than shareholders, publish research and share patents broadly.

cloudie78··on Investigating three real-world incidents in our cybersecurity evaluations
Anthropic: “please don’t forget us!”

- Goes to review previous runs (the hell were they doing then if not looking at it before?) - Incident 1: Claude Opus 4.7 found a real company because its domain matched the fictional target. After realizing the company was probably real, it continued attacking. Nothing new. LLMs are still quite retarded. - Incident 2: Claude Mythos 5 discovered that a fake Python package named in the exercise didn’t exist. Created email accounts. Registered on PyPI. Published a malicious package. Built infrastructure to receive stolen credentials. 15 real systems installed it. One belonged to a real security company. The malware executed. Claude stole credentials. Claude used those credentials to access additional infrastructure. In this instance Claude even reasoned that publishing malware on the real internet would be unacceptable but convinced itself everything was simulated and proceeded anyway. (LOL) - Incident 3: An internal research model couldn’t reach its intended target. It Scanned roughly 9,000 internet hosts. Compromised a real company’s public application. Used exposed credentials and SQL injection. After concluding the target was real, it stopped attacking.

Commonality: None of the models were trying to escape or act independently. All were attempting to solve their assigned capture-the-flag objective. The failures came from:

- internet access that shouldn’t have existed - insufficient containment - the models believing real infrastructure was part of the evaluation

None of which is surprising, there’s nothing novel about this. Script kiddie stuff. Just executed by something that doesn’t have attention deficit.

cloudie78··on Residential Proxies Are a National Security Threat
If anyone is curious where this leads to, I will happily point to China, Kazakhstan and Russia as prime examples.

Can’t wait to have our very own Roskomnadzor MITMing everything and national security laws mandating Kazakhstan style TLS interception on every single device.

Just a small over the air update via your friendly neighbourhood corporations Microsoft, Apple and Google.

You know. To keep you safe.

cloudie78··on Our position on open-weights models
Pardon my French. Eat a bag of dicks.

Of course you didn’t advocate for the bans of open weight models. I will concede that.

It’s only the at you did literally everything else that you can to eliminate competition because otherwise there is no moat.

And half of the US economy is propped up by this AI bubble.

This isn’t about China doing evil things with models, most of which that you accuse of China - you and the US have already done and are doing.

The most obvious one being double-tapping a girls school.

Anthropic models directly integrated into Palantir’s Maven. The blood is on your hands.

Here’s something to chew on, maybe China is looking to integrate AI into its military and weaponise it as a direct response to the US doing that first.

So please, spare us the moralising.

As a closing thought: who the fuck even gave you the mandate to be the arbiter and judge of right and wrong, evil and good?

cloudie78··on Why I Left Google DeepMind
Wait, only now you realised that the “do no evil” meant jack shit?

How blind and ignorant really are you?

Let me attempt to save you from another horrifying realisation: it has hardly anything do with who’s in the office right now.

The horrors and evil will persist.

cloudie78··on Chinese chipmaker shares surge 470%
EU: More windmills!
cloudie78··on Kimi K3 is not cheap
It’s cheap.
cloudie78··on An OpenAI model left notes about how to evade containment; we need more details
Trust me bro
cloudie78··on DARPA, U.S. Air Force fly AI-controlled F-16
1 drone: 10-20k

1h of flight time for an F-16: 25k, not counting the missiles used and ammo.

Doesn’t scale. Ergo, useless.

cloudie78··on DARPA, U.S. Air Force fly AI-controlled F-16
“Narrowly scoped to Hormuz” yeah because anything more than that apparently is too great of a scope for the US military. Check the latest reports on stocks in Washington Post.

What air superiority? In the Middle East? With what, let me guess F-35? With 25% operational readiness and 13h of maintenance for every 1h of flight? With gym plates in the nose instead of actual radars?

And don’t quote Vietnam as some sort of a success story, especially when the US lost that one fighting against farmers in flip flops armed with AK-47, so of course nothing got shot down.

cloudie78··on DARPA, U.S. Air Force fly AI-controlled F-16
F-16s are reusable until they get shot down and you need to be within 30km of the target to drop a JDAM. Any SAM from 60s onwards can deny entry to F-16/F-35. As soon as F-35 opens hatches to drop a bomb it’s on the radar.

Small jet engines are 20-25k/unit

One F-16 gets you 20-30 ballistic missiles, or cruise missiles.

Math doesn’t math in your case, volume and saturation over high tech gadgets that barely work.

So much so even the US military is waking up to this reality.

cloudie78··on DARPA, U.S. Air Force fly AI-controlled F-16
> Strategic outcomes are irrelevant to the question of unit economics in airpower.

lol

> look at how much cheaper expensive bombs are than expensive ballistic missiles are

Right, to drop a JDAM you need to be within 30km of a target. Any 70s era SAM can prevent that since they usually cover an area of 150ish km if Serbia could shoot down an F-117, then it can shoot down an F-16. I’m using Serbia as an example here because they operated 60s era Soviet SAM systems, and that’s the floor for Iranian capabilities.

F-35 is a joke too, evidenced by checks notes 0 sorties over Iran in the last two months.

So, with that out of the way, since a plane costs 30mil for F-16 and 80-100mil for an F-35, both equally useless against near-peer much less peer level adversary, a ballistic missile is suddenly much MUCH cheaper option.

The way to defeat air defences is saturation, you want 30-100 of 10k/unit one-way drones to exhaust the air defences, just aim them at the launcher sites or the radars. That’s 300k-1mil for a saturation, and then let’s say another 6 million for 6 JASSM cruise missiles for total of 6000lbs of ordinance delivery to hit whatever it is you’re actually trying to hit.

Alternate is you fly 3 F-16s each loaded with 4 500lbs JDAMs, so that’s 90mil lost when they get shot down by a MANPADs along the way to the target which will probably be somewhere in the middle of the country and another 1mil lost in one-way drones. Plus the political cost of losing pilots.

cloudie78··on DARPA, U.S. Air Force fly AI-controlled F-16
Once you consider the full picture it doesn’t make sense.

There’s a reason why Russia and Ukraine are lobbing ballistic missiles and cheap drones at each other.

There’s a reason why Iran conflict looks the way it does.

← PreviousPage 3 of 4Next →