mikewarot

Machining is a good skill to have. The ability to shape objects to tight tolerances comes in handy from time to time.

Personal anecdote: It was circa 1985 that I encountered a similar problem. I hand a hand held barcode scanner that had to upload the collected data via a proprietary serial protocol, for eventual reporting, to a PC.

The customer asked "what happens if I disconnect this right now?" during the upload. It was a corner case I hadn't considered. It only took a few days to make it bulletproof.

---

The authors don't seem to consider the case of very slow or intermittent Internet.

I'd be worried about corruption in the download, and check for that. I'd also make very sure the watchdog hardware was on and the code respected it.

It seems to me you need 3 buffers for the OTA code, not 2. There should be a way to keep the last version that runs for X seconds in addition to any new updates, where X is quite large, perhaps a whole day.

I remember 38400, 8, n, 1

38400 bits per second, 8 bits per character, no parity, 1 stop bit

All bleeding stops.

If Ed Zitron is to be believed, Anthropic, and all the other LLM based tools companies are bleeding cash at one of the highest burn rates in history, subsidizing your tools, and likely giving them to you at around 5-10% of their costs. Even if you're prepared to pay 10x or more for the same thing, most aren't, and they don't have a sustainable future.

Your best bet is to develop a local capability on your own hardware while you've still got Claude to compare it to.

Second best is to have an LLM model and harness you can run in the cloud independent of Anthropic or any other AI company.

>Now imagine that we had spent all these resources on doing the basics: ensuring your organization has an up-to-date and comprehensive asset inventory with fine-grained package listings; building infrastructure that supports regular, frequent, and automated OS and applіcation updates; automatically rebooting systems when they hit, say, 30 days of uptime to ensure these updates are picked up;

That would still leave the root cause of most of our problem un-addressed, ambient authority based Operating Systems are fundamentally insecure. Unfortunately, all of our systems share that architecture.

To answer the question as I misread it:

  What's the most career altering *course of action* you've gone through?

I was an IT guy, and I took a job making gears in a job shop. The pay and commute were horrible, but it was immensely satisfying. Some of the gears I helped make will still be in service in a century.

Uptimes used to be measured in years. Clearly we no longer have OSs or server software that are fit for purpose.

I think this could be the thing that finally pops the AI circular financing bubble, leading to World Depression II. I'm not sure either of the existing 2 faces of the duopoly infesting politics will survive, and there's a non-zero chance this breaks civilization as supply chains collapse, and we enter a new dark ages.

The Boy's Second book of Electronics by Alfred Morgan(1957) introduced me to electronics circa 1975, and lead to a technical mindset and lifestyle.

The containerizing is almost a capability based security system. Instead of adding that layer to block everything except a certain resource, why not change the defaults in the OS so that it defaults to NO access instead? Saves resources, provides much better security.

First you need an Operating System that can withstand direct Internet exposure for a decade without incident. Any administration required is going to kill it dead. There are no suitable OS options available at present.

Then you need to make the actual service dead simple to use from anywhere. Punching through NAT and firewalls are the main issues there.

The Memex as described in 1945, especially with the ability to recall and spit out a complete copy of a given path through Hypertext with all the references. It recorded everything, unlike our web browsers.

A capabilities based Operating System that could run Linux, Win32 and other OSs applications natively, replacing the system dialog boxes with power boxes.

An open data diode based on a pair of Raspberry Pi (without wireless of any kind, obviously) and a high-speed optocoupler.

A Streaming variant of SQL, so you could get a live view of any query that publishes updates

A variant of Lazarus, with much better documentation, and a new, magical equals in addition to all the existing code, like that from Metamine. You do the assignment, and it keeps the equality updated as the value on the left changes, like auto-recalc in a spreadsheet.

The true horror show is in IT administration in these companies. Haven't they ever heard of air gapped networks? Almost everything that happened was due to incompetent IT staff.

Air gap the supposed ASI, monitor/log it though a data diode, and you don't have to worry about egress of control.

It's important to know that there were multiple standard formats for floppy disks, with 40 and 80 tracks per inch, thus actually different track widths, and different data rates and encoding formats. Matching both is the key to getting a disk to read successfully.

However, you can, apparently, play the game in your browser here: https://www.myabandonware.com/game/super-solvers-spellbound-...

My career in software development happened back in the days of MS-DOS, before Windows really took hold. I developed a system that recorded the inspection of Fire Safety equipment like Fire Extinguishers, Hoses, etc. My initial estimate for the time required to build the system was 2 months, it actually took 3. Then the client looked at it, and it met all the requirements, but was unusable in practice. I then worked onsite for a year, and learned so much about software usability by seeing it in use. After that year, we deployed it to 10 other sites.

Other than having a summer intern once, it was a solo effort. The code was written in Turbo Pascal, with some code on the hand-held computers written in a language called PL/N. (A variant of PL/1 for the Norand corporation handhelds)

I wrote all my own text editing/forms handling libraries, and everything else I needed. Version control was using pkZip to archive the source code onto 3.5" floppy disks.

Tying it back to the article in question, I can't imagine what it's been like with the push for "sprints" and all the rest in software development over the past few decades. I can see how it would drive you to burnout. I'm very thankful I missed it.

Y'all have my sympathy.

The key feature of a flip phone is that when it's closed, you don't get random user input. My cheap Motorola Power G 2024 seems to like to open random apps when transiting to/from my pocket.

I also miss keyboards, typing on glass really sucks.

That's certainly handy... here's a simple audio frequency shifter I threw together[1].

First it takes the audio from your microphone. It then feeds it through a Hilbert transform to convert incoming audio into an IQ (complex) format. The advantage of an IQ signal format is that when you multiply it by another signal, the frequencies add or subtract, without getting extra image frequencies. This is why it's worth doing twice the computation in DSP work.

Then your audio is multiplied by a cosine wave with a frequency determined by a GUI slider, from -2000 to 2000 Hz.

This has the net effect of raising or lowering the pitch of your voice, like tuning in a single sideband transmission on ham radio.

To get the results out to the speakers, it samples only the real component to send to the audio output.

The net effect is it can raise or lower the frequency of your voice.

[1] https://gnuradioworld.com/#fg=H4sIAAAAAAAACsVXS2_bOBC--1fMTR...

Nowhere in any of the scenarios they let you model, do they consider the cost of computation for inference, nor training of new models. It may be that after the subsidies can no longer be offered, most AI use becomes cost prohibitive.

This would lead to a much, much smaller "total addressable market", rendering the massive AI build out into a rapidly depreciating asset that can't even be operated without a loss.

Even if the hardware was effectively free following a market crash, and bankruptcy sale, the power required to run it might be more expensive than the compute it generates is worth.

At which point, unlike dark fiber after the .com crash, all the hardware becomes scrap metal.

Now factor in the macro picture, in which this AI bubble is holding up the stock market, and by proxy, the US economy. I think we're in for Global Depression II, worse than 1929.

I regret that we don't have sufficiently strong operating systems. I don't know what I could have done differently over the past 40 years to have changed that, I doubt there is anything. This is, in my estimation, the root cause of much of the grief we now suffer with the Internet in general.

Because we can't run our own servers without substantial risks, almost everyone is forced into the walled gardens where network effects make it feasible to economically run the servers, and the administrative staff to keep them running, and somewhat secure. This provides the honeypot that then draws in advertisement and all the ills that go with it, as we're not the customers, we're the product.

I would hope that we can collectively get our security act together, which will as a byproduct make it easy for everyone to run their own servers without risk. But that's always been 20 years off, like fusion power. Everyone seems to think that Linux is good enough, which it isn't, but here we are. 8(

> leverage AI to implement verifiable enforcement of the invariants

The only verifiable invariants that are going to work in the long run are:

  air gaps
  data diodes

Interposing relays used to be used with PLCs[1] to prevent motor control systems from energizing both forward and reverse power to a large motor in the unlikely but possible event of control systems failure. We need to apply the same amount of engineering rigor to our computer systems.

We simply shouldn't trust software to a job that belongs in hardware.

[1] https://en.wikipedia.org/wiki/Programmable_logic_controller

Not radical enough, there are only 3 mentions of "air gap" and zero mentions of data diodes in the article.

The focus seems to be more of the same failed strategy:

  a shift to "safe" programming languages
  a focus on patching known bugs proactively
  ratcheting up rules and enforcement

This all amounts to "try harder", which isn't going to work. We need a ground up restructuring of everything if we want to have a stable foundation upon which to keep society going.

If you run critical infrastructure, it must be air-gapped from the internet. If you need to monitor said infrastructure, there are data diodes which can be configured to allow the egress of monitoring data, with ZERO risk of allowing ingress of control, enforced by the laws of physics, not some code running on hardware that might not be perfect.

We need to channel everything important through clearly defined, and well monitored channels.

Ambient authority based operating systems were a great hack, but we need to leave them behind. Proven microkernel based OSs with everything running in userland, and all access provided using capabilities, with UI including powerboxes, are a bare minimum.

The amount of code and hardware we actually trust should be pushed towards zero as far as is possible. Linux, Windows, etc are all unfit for the present and future security needs of humanity.

As for AI/LLMs themselves. They're natively a huge amount of floating point math that results in token predictions. The code that runs the math is something we can control. There's no reason to give code access to anything about the machine it's actually running on. This should be something that can be enforced without extraordinary effort.

This is the result of a bi-partisan effort[0], to divide the country, and keep our focus away from the concerns of the donor class.

First, Ronald Reagan's adviser Lee Atwater[1], advocated for the use of wedge issues, such as the mythical "Welfare Queen"[2], to erode support for the social safety net. Reagan's FCC commission decided end the Fairness Doctrine[3], which helped maintain the public consensus in the mainstream. In my strong opinion, it was a huge mistake, along with Trickle down economics. However, considering the traitorous nature of Reagan's administration[4,5], I guess it's to be expected.

Hillary and Bill Clinton went along with things, and came up with the strategy of "Triangulation"[6], in which they used wedge issues as a way of staking out a political position, without actually needing to care for their base. This resulted in a wave of virtue signaling[7] and political correctness[8], and an end to actual support of the social safety net. Examples of this include "Don't Ask, Don't Tell" as official policy in the US Military.[9]

When Hillary Clinton and Debbie Wasserman-Schultz rigged the 2016 Democratic Primary[10], they shunted the backlash against all of this that Bernie Sanders was channeling, and gave us Donald Trump, who is a master of exploiting it.

[0] https://en.wikipedia.org/wiki/Wedge_strategy

[1] https://en.wikipedia.org/wiki/Lee_Atwater

[2] https://en.wikipedia.org/wiki/Welfare_queen

[3] https://en.wikipedia.org/wiki/Fairness_doctrine#Revocation

[4] https://en.wikipedia.org/wiki/1980_October_Surprise_theory

[5] https://en.wikipedia.org/wiki/Trickle-down_economics

[6] https://en.wikipedia.org/wiki/Triangulation_(politics)

[7] https://en.wikipedia.org/wiki/Virtue_signalling

[8] https://en.wikipedia.org/wiki/Political_correctness

[9] https://en.wikipedia.org/wiki/Don%27t_ask,_don%27t_tell

[10] https://www.politico.com/magazine/story/2017/11/02/clinton-b...

>it's important to note that

Here on HN, I've noticed that I pro-actively defend myself against things I think will be said in counter to my comments. I concluded it's being trained into me.

Alternatively, I've been banging on against the same things that I see as insane so long that I know all the objections.

Relevant(ish) XKCD: https://xkcd.com/3290/

Why the heck don't they air-gap these things when they're testing them? The can download the models and training data from read-only sources, and use a data-diode to a allow external monitoring without risking egress of control.

It's not rocket surgery!

I get that the Project 2025 is an evil plot, but that's missing the existing duopoly that allows the Donor class, and their alliances, both foreign and domestic, to override the will of the people.

Even if every single Republican got voted out, the rot is still there, just wearing the other hat.

I'm not sure which he replaced, but having started trying to grok how it actually works, it's chock full of HFET and other amplifier chips, and over-stressed RF switches.

I'll report back if we ever get it actually working properly.

The incentives have disappeared. I'm old (64), and when I started out, there were printed trade publications for pretty much any discipline, that showed up in the mail, paid for by advertisers. The articles were deep and useful, because they had to respect the audiences time enough to avoid the trash bin.

The money that funded those got sucked up by ad-tech companies like Google and Facebook.

Blogging, was a low to non-profit version of the same thing. You'd direct your RSS reader to your favorite feeds, read what your "friends" wrote, and reply in the comments, etc.

The thing that happened was the audience shifted from technical people happy to deal with RSS readers, and do the work of maintaining subscriptions found the process of getting a blog going to be burdensome, and moved into the easier walled gardens.

The effort of blogging to a stand alone platform wasn't horrible, nor was writing and editing something long enough to be worth the time and consideration of the reader. To make it worth it for a blogger though, you need an audience, and the feedback cycle was slow, through Google for anyone not using an RSS reader. It was really hard to get attention enough to feel the effort was worth it.

The social networks arrived, and they offered faster feedback, and more of it, with a single mechanism to handle replies, instead of each platform having a different system, and anti-spam mechanism. When Robert Scoble gave up on blogging, that signaled the end of it for me.

A technical blog is a teaching tool, and nobody is willing to pay the teachers enough in money, or attention, to make it worth it.

One area that has bloomed though, because that's where the incentives moved, is video blogging. Here are some YouTube channels that might be interesting to you:

https://www.youtube.com/@HuygensOptics -- if you want to learn about optics

https://www.youtube.com/@EEVblog -- an opinionated Australian with one of the oldest EE focused channels

https://www.youtube.com/@CuriousMarc -- deep technical/repair videos

https://www.youtube.com/@ThinkingandTinkering -- alas, Robert Murray-Smith is no longer with us, he had many interesting things to say

https://www.youtube.com/@ThisOldTony -- Machining videos with a sense of humor

https://www.youtube.com/@smartereveryday -- an engineer who used to field test missiles learns new things all the time

https://www.youtube.com/@Blondihacks -- Canadian machinist who teaches along the way doing projects

https://www.youtube.com/@TheAmpHour -- an audio blog about electronics design

https://www.youtube.com/@Computerphile -- various topics related to computers with staff mostly from the University of Nottingham

https://www.youtube.com/@primitivetechnology9550 -- John Plant teaches by showing, how to build the most basic technology ever, he started with a rock and a stick, and worked his way up from there. Turn on subtitles, he doesn't talk, he annotates on screen, it's wonderfully quiet

https://www.youtube.com/@OCPTAP -- if you're a time nut, this is the channel for you, it's about very, very precise timing

https://www.youtube.com/@NileRed -- crazy and almost safe chemistry experiments

https://www.youtube.com/@urituchmanpigeon -- an eclectic approach to building things

https://www.youtube.com/@CaseyJones-Engineer -- an actual engineer who picks apart various things gone wrong with various current events.

https://www.youtube.com/@mrpete222 -- Mr Pete, retired shop teacher with an attitude, and a lot of things to show you how to do.

https://www.youtube.com/@PracticalEngineeringChannel -- Grady, another actual engineer, shows how infrastructure works, with practical demonstrations built in his garage, and some on-site things as well

https://www.youtube.com/@DavesGarage -- Dave Plummer, the guy who wrote task manager, and the zip file handler for Windows, talks about various things computer related

https://www.youtube.com/@AnastasiInTech -- A chip designer talks about the latest trends in chip fabrication

https://www.youtube.com/@JeffGeerling -- Things related to radio broadcasting, raspberry pi, mini-racks, gps timing, etc. His dad is a broadcast engineer.

https://www.youtube.com/@RealEngineering -- What it says on the time, breakdowns of how things actually work

https://www.youtube.com/@wgowshipping -- Sal Mercogliano, an actual mariner and Professor, discusses current events in shipping on the high seas.

https://www.youtube.com/@MentourNow -- A pilot discusses current aviation events

https://www.youtube.com/@NDC -- NDC Conferences - Interesting programming talks

https://www.youtube.com/@ZackFreedman -- Wild and crazy, the inventor of GridFinity

https://www.youtube.com/@voiceofthunder9620 -- An actual scientist, a chemist, who really, really despises Elon Musk, and sometimes plays with NaK, a liquid metal at room temperature. He might be the one who gives us a work-around for global warming

https://www.youtube.com/@NoEngineerHere -- Interesting projects, but NOT an engineer

https://www.youtube.com/@StefanGotteswinter -- a proper German doing machining projects

https://www.youtube.com/@dgelbart - Dan invented the first practical stepper, which gets used in the manufacturing of every chip ever. You really should watch his "how to make things" tutorial

https://www.youtube.com/@3blue1brown -- Grant Sanderson talks about Math, beautifully illustrated and story-told

https://www.youtube.com/@Kathy_Loves_Physics -- The real history of how physics got to where it is

https://www.youtube.com/@oxtoolco -- Recently retired from Berkeley's University machine shop, lots of practical tips related to machining very precisely

https://www.youtube.com/@TITANSofCNC -- CNC manufacturing, they offer free education in the trade

https://www.youtube.com/@Clough42 -- Machining and CNC related projects

https://www.youtube.com/@arduinoversusevil2025 -- Arduino vs Evil, an opinionated Canadian who swears a bit does various things related to machining, with a lot of tear downs of tools.

https://www.youtube.com/@WardCarroll -- Former Navy Aviator talks about current events, sometimes with Sal from What's up with Shipping (see above)

https://www.youtube.com/@tested -- Adam Savage, of Mythbusters does various builds

https://www.youtube.com/@w2aew -- Radio theory and projects

https://www.youtube.com/@AppliedScience -- Some of the coolest science projects, including things like spin coating, supercritical CO2, etc.

https://www.youtube.com/@RexKrueger -- Traditional Woodworking

https://www.youtube.com/@DylanBeattie -- Dylan is great at breaking down things that you think are simple... like "Plain Text" formatting

https://www.youtube.com/@ShakeTheFuture -- Melting metal in a microwave, so you can do home casting

https://www.youtube.com/@joepie221 -- A machinist in Austin Texas, with a lot to teach

https://www.youtube.com/@KevlinHenney -- Talks about programming

https://www.youtube.com/@GOTO- -- GOTO Conferences, programming talks

https://www.youtube.com/@essentialcraftsman -- Practical construction and blacksmithing

My late friend Ward was a strong supporter of Build-a-Blinkie[1] and their mission to teach people to solder by doing.

My friend Jim's been soldering since the 1950s. We solder a lot of things with leads, for this, his method is to get each item to be joined tinned, then use solder as a mechanical joint. Melt in the solder, remove the iron but hold them together, then once cool, tug on the joint to make sure it holds.

[1] https://build-a-blinkie.org/events.php

There are always insane people waiting with a plan, its the job of government, and politicians to keep them at bay. [1,2]

[1] https://en.wikipedia.org/wiki/Business_Plot

[2] https://en.wikipedia.org/wiki/Operation_Northwoods