Software engineer.
Twenty years in, still diggin' it.
Code is wizardry.
Blu-ray archivist.
Mostly cerebral cinema.
Lots of good shows, too.
Pinball enjoyer.
Would like a machine of my own.
Saving up for one.
Original music: https://twuni.net/
GitHub: https://github.com/canterberry
I don't want to outsource my thinking to a machine.
I don't want "art" averaged from some aggregation of actual artists' work.
I can write my own fucking emails.
I can write my own fucking code and design my own fucking games.
And I want art that comes from the brains of human artists. Who feel and whose soul goes into their work.
All this stuff is clever and I am sure there are uses for it but the things that are actually valuable in life can't be reduced to a fucking prompt.
FFS. 🦬
Reading the program and talk titles/synopses of some AI design uni symposiums. Many are sounding a little blue-eyed and I wish at each event at least one of the talks would also involve a more holistic, critical Realpolitik dissemination of AI use and address glaring contradictions (e.g. see the recent talk by Goldsmith's @danmcquillan for example), even if it's hard to hear for some...
Designing for more inclusiveness & empathy, incl. for "non-human" participants sounds great & long overdue as a design (and education) focus, but does it really require (or even should require) any AI involvement at all? Does it make any sense at all, using the most centralized, monopolistic, resource intensive, extractive/abusive and environmentally/socially hazardous form of computing to help contemporary Design Practice to become more inclusive/empathetic?
Is it not a total conundrum to even talk about any "post-extraction" aspects in LLM-based AI, if the entire conceptual foundation and actual implementation rests on (often illegal) extraction of all forms of knowledge and physical resources to ensure its continued growth/scale/relevance? Where does the data, the minerals, energy and water come from for the build-out of the intended capacity? At what costs?
Does it make any sense to talk about speculative AI design utopias, but at the same time base all the routes/solutions to get there on funding/using orgs who're main culprits/contributors to the current dire state of global affairs, and who're continuously abusing their position and pushing for more erosion of existing legal frameworks, for more surveillance (to increase data intake and build out monopolies) and dissolving political/environmental regulations/protections to increase their extractive practices?
Does any of this really empower humans (rather than individual people/groups involved in AI proliferation) or does it objectively improve the situation for any other _living_ organisms on this planet? Not talking about AGI threats here — the current set of factors is more than sufficient — isn't increased use of soon hyper-scale AI one of the biggest risks? Is there a talk at any of these events about cost/benefit analyses and also an overview of which parties/groups/demographics stand to cost and benefit?
Will AI help to solve inequality or isn't it (becoming) part of the cause? How will governments respond to massive job losses, resulting loss of consumers/markets, increased chances for social unrest, coupled with increased energy prices, inflation and technological possibilities (and active offers by suppliers) for increased surveillance/enforcement? Is there any institutional research on useful design approaches for helping people in any of these AI-induced situations?
How will AI preserve people's autonomy of personal computing if more and more infrastructure becomes centralized/surveilled/censored and people without the latest hardware become excluded from state-provided services? How can we trust any AI proposed design solutions/approaches with their more-than-shaky epistemological grounds, lack of rigor/provenance, using possibly invisibly hostile/toxic ideas/philosophies and the generally stochastic approach to generating non-reproducible "answers"? What additional design processes are required to make any of this actual practically usable, also in light of legal requirements/certifications in many fields?
If the framing of "more-than-human design" is going to be about "AI empathy", then we're entering another very dangerous territory, even if this all falls under Speculative Design.
In 2025, it's about time to get real and each time it's a missed opportunity to continue treating SD in a vacuum, entirely disconnected to our current timeline, much like it was done a decade or more ago, and much like how architects still keep on dreaming up vapid design utopias, kindly sponsored by some of the most human-rights-abusing governments on the planet...
Startups are prime hacker targets due to weak defenses; post-breach costs (forensics, legal, compliance) often exceed hundreds of thousands, leaving investors to absorb the damage. #CyberSecurity #Ransomware https://www.wsj.com/articles/deep-pocketed-investors-make-startups-a-target-for-hackers-ab4e0f9a
#GetFediHired #Seattle #RemoteWork
Well, it's looking like the job I started back in December was a swing and a miss. I'm grateful for the investment on their end, and am bummed that I wasn't able to really shine in my role there.
In case you know of something, here's a brief intro:
I'm a seasoned full-stack #SoftwareEngineer, and I've got what you might call a unique set of skills.
I'm a hyper-minimalist and love to build tightly scoped, lean software that does one job exceptionally well using as few dependencies and resources as possible.
I build web client apps that don't require any dev tools or build process. They're fast, accessible, fully localized, responsive, and you can even install them as mobile apps if you want.
I build scalable systems using stripped-down variants of Kubernetes and built-for-purpose Helm charts.
I provision cloud infrastructure with just the right amount of cost and observability to know when and how to scale horizontally or vertically.
I cover everything with comprehensive tests.
I continuously deploy all the things, with just the right amount of process to minimize risk and maximize delivery velocity.
I've built entire SaaS products from scratch, several times over. Serverless app deployment via blockchain; fully customizable product demo templates with live chat; a precision ultra-refrigerator for your fresh produce; just to name a few.
I wrote my own streaming video platform that I actively and regularly use.
I've got over 200 projects on GitHub and GitLab, and that's just the public ones. Some of them are even kinda popular and useful.
I'm looking to join a company where I can really shine -- either by doing things like what I've described above, and/or by empowering teammates who want to learn and follow my lead.
Actually screw it, here's the funny thing I was going to post, before I decided to post about the EU/Trump thing:
In the UK, we've had internet censorship for years. In particular, torrent sites and many streaming sites are blocked by court order on major ISPs.
However, ISPs report what they block E.g.
https://www.virginmediabusiness.co.uk/help-and-advice/technical-support/list-of-court-orders/
https://www.virginmedia.com/help/list-of-court-orders
Here's the funny:
They are the best places to find what sites are available. Better than TorrentFreak. Then just get a VPN, or uncensored ISP.
Dad and I built a guitar. A Les Paul single cut. We sanded, painted, glued and sweated in the heat of a Bermudian summer to make it happen. We fixed a lot of broken things that summer. That guitar still screams in the attached song I covered 2 years ago. And today its getting played. RIP
Okay, so ranking the top artists I have paid (indirectly) via Spotify streams, the top artist is...
...Aqua Teen Hunger Force.
Because I listen to their Christmas album more than I care to admit.
Apparently, my streams got them paid $0.07.
There are about 70 more artists I've listened to enough on there which resulted in those artists getting paid, and many more that I guess didn't get paid at all. All 70'ish of those were paid less than $0.07.
Spotify, as of today, costs $11.99/month.
My streams, over all time, have collectively gotten artists paid around $0.70.
Something does not add up here.
When Signal was designed, our threat model was protecting the communications of civil society, journalists, just regular citizens ...
The threat model of military operations & sharing your hate of Europeans was not what Signal was designed for. Ephemeral messages and cryptographic deniability are not fit for communications that require accountability.
But I appreciate their effort to make government more efficient by adding journalists to the chat instead of requiring to go through FOIA.
After years of using #DuckDuckGo as my primary search engine, I'm finally letting go and exploring other options for my go-to.
Lately, the search results have just been swamped with bullshit AI generated websites. Even for basic searches, the entire first page or two is this. It's become completely untenable.
Plus, virtually every news result is wrapped in MSN. So I stopped using it for news a long time ago.
Also, I've been noticing "helpful" no-click AI summaries on searches, lately. 😬
What I'd love to see is a privacy-preserving, #Tor friendly, JavaScript-optional, search engine of community-vetted results.
Open to suggestions and recommendations!
Wow.
With #Xfinity, it is no longer possible to talk to a human.
There is no customer support phone number listed.
The old support number still works, but all it does is direct you to the Xfinity AI Assistant.
The online account management interface doesn't even work for doing basic things. It times out or outright fails with server errors.
No way to cancel service, add a payment method, change plan, or anything at all.
And the AI assistant just links to the web interface that's broken. It also doesn't respond to requests to speak with a human.
This is the world we live in.
Today's code performance quote:
"When you treat the computer like a magic black box that takes inputs and produces outputs, you get slow code"
I just deleted my book.
A thread about choices, and doing the right thing.
Because we're all facing a lot of that right now.
🧵
> The basic idea is to turn the American government into the biggest crypto bag-holder of all time. If the plan goes through, hundreds of billions of dollars of public assets will be spent or leveraged to buy a million Bitcoins
https://prospect.org/power/2024-11-26-crypto-plot-against-americas-gold-reserves/
Oh wow, the cryptocurrency people must be absolutely furious with Big Bad Government so clearly interfering with their libertarian dream of truly independent currency!
They're furious, right?.. 👀
Today the Trump Administration terminated the federal government's only advisory committee on open government. GovTrack's founder was a member of the committee.
https://joshuatauberer.medium.com/trump-kills-nations-first-open-government-federal-advisory-committee-firing-me-along-with-it-258dc0971afe
LET'S TRACK THE WHITE HOUSE
We're expanding our mission if we can get enough subscribers for a new newsletter tracking executive orders, DOGE, and other major official actions coming from the President and their appointees.
Just as we do for Congress, we’ll cut through the bravado and hyperbole: This is a project about what your government is actually doing.
Plus, you can get a t-shirt.
Please back our project and share!
https://www.kickstarter.com/projects/govtrack/alert-new-executive-order?ref=8f0vhi
Have fun!
https://gulfof.mapquest.com
Our "2024 Report Cards" for legislators are now posted. These are our biennial statistics on cosponsorship, missed votes, bipartisan-ness (not that that really matters anymore), most laws enacted, etc.
Higher numbers isn't better. At least, that's not for us to say. The point of all this is to get a deeper sense of how Congress operates and how your legislators conduct their work.
I have started browsing the news exclusively via Tor with JavaScript disabled.
A+. Highly recommend.
You'll wonder why you ever did it any other way.
