Jon Lowy

Jon Lowy

@jonny-spanner.bsky.social

I'm an engineer, writer, 11.4% equipped for life. Puckish and prickly. Handle with care, contains traces of nuts. I'll always follow. Unless you're porn or coin. Return the love? KEEP YOUR SHITCOIN TO YOURSELF. NO TRADING/INVESTING ADVICE PLZ

12.7K followers28.7K following4.4K posts

Jon Lowy reposted

My heart is heavy. Walter Koenig, our Ensign Pavel Chekov, passed away early today. Walter and I met on the bridge of the Enterprise in the 1960s, and we stayed friends for nearly six decades. When Brad and I married in 2008, Walter stood beside us as our best man.

Not sure anyone is interested, but I've been trying to develop anninage related workflow to make my CAD outputs look better without so much effort. AI is not taking anyone's job over any time soon

If this is indeed the opening act of the next world war-proper, its going to be fought by infantry columns on long marches - and drones with solar panels

Again time for apostacy. I've said before, I repeat here. I'm an engineer. If a thing is a thing, then its a measurable thing. I'm not commenting on climate change, or AGW. I'm just observing that catastrophizing bad weather is actual bullshit. Show the data extremeweather.thehonestbroker.org

extremeweather.thehonestbroker.orgUS Extreme Weather and Climate Change Dashboard — Trend AnalysisA dashboard tracking 38 US variables across 8 extreme-weather phenomena, detecting long-term trends against official NCEI/SPC/USGS/USDM/MTBS/NIFC data following the IPCC's own detection framework.

I stopped once I was getting potty-worded and the tool "apologized" for basically being unable to follow instructions and assuming I didn't GAF about the details. Autocomplete remains an idiot savant with no knowledge or ability, just a lookup table for average responses/completions. 2 of some

View full thread

Sorry for the excess words. Do not believe a WORD of the ROI claims for the IPOs. They will never return a profit selling LLM auto finish for complex tasks. Your job is safe, because they cannot EVER teach an LLM fish to ride a real-job bicycle.

Also forgive the fat finger typing. Small screen, fat fingers old eyes, a hugely dyslexic professional writer. A fish that learned to ride a bicycle - though not with precision.

View full thread

My leap was a trial for illustrations. I wanted to see if meme creation meant useful illustration skills. Not to use the output, but hoping to use it as inspiration. Nah. Once the illustration is required to be *useful/informative* it ain't. 5 of some

So the next step was to see if it could follow very clear instructions in generating CAD models. With thongs to copy as a guide. Nope. It all cones down to purpose. Imposing purpose through prompts TAKES LONGER THAN DOING THE JOB. Conclusion: LLMS can only do simple stuff (kinda) OK. 6 of some

View full thread

I even got paid (surprisingly well) to so A/B/C selection "training" on slop, by a client I shall not name, who was "developing" their own content slop maker. So I have good operational experience with Chat GTP, Claude, and recently Googles tool. 3 of some

That does not make me an expert, other than knowing a bit more at the coal face than most. So I took the leap, I've paid OpenAI a tiny bit of money to use their tool. I don't want its written output. As a former teacher, I'd grade it a C- on a good day. And it has few good days in my field 4 of

I watched the "AI can do this" wave arrive and create a massive panic attack the top end, every client desperately using every checker they could find to chase diwn tge cheat writers. I even developed a skillset - taking the mediocre AI output and twisting it to maje it human. 2 of some

I wun 2 chaire sum AI appo-ca-lips funking, Cos I bin dun sum thawts. (No, an AI would be unlikely to read that sentence, yet its clear to you.....kinda my point) If you're concerned about AI heaving you our of your job...... (1 of not sure how many)

View full thread

Ok, i hear you say, that's not simple enough to show security in most jobs. I assert otherwise. Excel may look complex to a call center operator, a restaurant worker, a data entry operator. It's not. It's a bell weather. You don't autocomplete in your job - even if autocomplete might help you.

The AI bubble is making fortunes, and the lies it tells to make sales are the usual bullshit. The emperor (as with most such) has no pants on. This too shall pass. Yes, real artificial intelligence will be great, or kill us all. But ain't nobody making that paradigm shift with LLM autocomplete.

I wun 2 chaire sum AI appo-ca-lips funking, Cos I bin dun sum thawts. (No, an AI would be unlikely to read that sentence, yet its clear to you.....kinda my point) If you're concerned about AI heaving you our of your job...... (1 of not sure how many)

View full thread

This layest one has crystallized just as AI promised to end all jobs and create leisure (starvation) for all. There is almost no part of any of my careers that AI can do more than screw up. AI is very high grade, really capable autocomplete. Not intelligence. (4)

It can't even write a good quality blog. If your job EVER INVOLVES DECISION MAKING OR IMAGINATION, your manager may not realize it yet, but a token based agent will simply bollox it up. It can't even write a basic excel routine, without endless hand-holding, depite a rules based process. (5)

I wun 2 chaire sum AI appo-ca-lips funking, Cos I bin dun sum thawts. (No, an AI would be unlikely to read that sentence, yet its clear to you.....kinda my point) If you're concerned about AI heaving you our of your job...... (1 of not sure how many)

Back story. I've had a non-conforming life. I've had careers as: - a robotics engineer - a high school teacher - a materials scientist - an inventor - an investment manager - an R&D team leader - a founder and funder - an oilfield engineer - a family office principle - a CEO (2)

So i can speak to my own jobs (many and varied), and the jobs of hundreds of people who have worked alongside me, or under my direction. My (not last) career is as a professional writer - high end web site engineering and tech content. About 6 million words over 5 years. (3)

Question: Among all the moral and political outrage, has the alt. media and discussion missed the point of the Iran "excursion" entirely? Postulate: The death, destruction, and economic catastrophe that follows IS THE PURPOSE. 1st

How do you move the US towards a more passive response to the swing to fascism? Distract. Suggestion: Stop leading on moral terms and focus on the strategic implications. No amount of outrage will influence this. Only exposure of the subtexts can shine useful light. Follow the money.

Who profits from oil price chaos? Who shorted stocks recently and when? Which sanctions of secret friends are being eased? Who owns the benefits that these real-estate players and data brokers are riding? Forget morality. Outrage is irrelevant, a

You know how sometimes you dark suspicions about the state of being, click into the evidence you didn't see clearly, until it was written in sky writing? A NEW, New Deal required. Why neo-con, neo-lib have created systemic failure on a continental scale: thedailyeconomy.org/article/esca...

thedailyeconomy.orgEscape Velocity: Why America’s 1963 Poverty Math Is Broken | The Daily EconomyThe Kennedy-era poverty formula hides the economic pressures now crushing working Americans. Worse, benefits losses are wiping out wage gains.

Jon Lowy reposted

ATTN: If you’re one of the 1,000 FBI agents who reviewed the Epstein files, feel free to reach out to me via signal (in my bio). I’ll keep you anonymous- as I have the thousands of current and former federal workers who have sent me tips. Please share this widely. 🙏🏻