Jon Randy
LightNews LightNews
banner
jonrandy.bsky.social
Jon Randy
@jonrandy.bsky.social
37 followers 46 following 99 posts
Posts Replies Media Videos
jonrandy.bsky.social
"AI is the asbestos in the walls of our technological society, stuffed there with wild abandon by a finance sector and tech monopolists run amok. We will be excavating it for a generation or more."

doctorow.medium.com/https-plural...

#ai #llm #chatgpt #openai
The Reverse-Centaur’s Guide to Criticizing AI | by Cory Doctorow | Dec, 2025 | Medium
My speech for U Washington’s Neuroscience, AI and Society lecture series.
doctorow.medium.com
December 6, 2025 at 5:36 AM
jonrandy.bsky.social
"LLMs are simply tools that emulate the communicative function of language, not the separate and distinct cognitive process of thinking and reasoning,"

www.theverge.com/ai-artificia...

#llm #ai #chatGPT
The AI boom is based on a fundamental mistake | The Verge
The AI boom is based on a fundamental mistake.
www.theverge.com
November 26, 2025 at 2:53 AM
jonrandy.bsky.social
"...this “training data” is far from the sum total of human knowledge, with oral cultures and even languages underrepresented or absent."

www.theguardian.com/news/2025/no...

#ai #llm #chatgpt
What AI doesn’t know: we could be creating a global ‘knowledge collapse’ | Artificial intelligence (AI) | The Guardian
The long read: As GenAI becomes the primary way to find information, local and traditional wisdom is being lost. And we are only beginning to realise what we’re missing
www.theguardian.com
November 24, 2025 at 3:23 AM
jonrandy.bsky.social
"Writing code used to mean leaving a trace of yourself behind — your taste, your discipline, your thought process. Now, it feels more like curating outputs from someone else’s brain."

Couldn't agree more.

medium.com/design-bootc...

#ai #llm #programming
The Lost Beauty of Code. When code stops being written by… | by Alessandro Romano | Bootcamp | Oct, 2025 | Medium
When code stops being written by humans, does beauty still matter?
medium.com
October 31, 2025 at 1:42 AM
jonrandy.bsky.social
"Four years later, he’s betting his company on its ability to sell ads against AI slop and computer-generated pornography. Don’t be distracted by the hype. This shift matters."

calnewport.com/is-sora-the-...

#ai #sora #llm #chatgpt #openai
Is Sora the Beginning of the End for OpenAI? - Cal Newport
On ​my podcast this week​, I took a closer look at OpenAI’s new video generation model, ​Sora 2​, which can turn simple text descriptions into ... Read more
calnewport.com
October 22, 2025 at 8:40 AM
jonrandy.bsky.social
"The quality crisis started years before ChatGPT existed. AI just weaponized existing incompetence."

techtrenches.substack.com/p/the-great-...

#development #ai #llm #chatgpt
The Great Software Quality Collapse: How We Normalized Catastrophe
The Apple Calculator leaked 32GB of RAM.
techtrenches.substack.com
October 15, 2025 at 2:21 AM
jonrandy.bsky.social
This.

theoatmeal.com/comics/ai_art

#aiart #llm #ai
A cartoonist's review of AI art - The Oatmeal
This is a comic about AI art.
theoatmeal.com
October 10, 2025 at 4:54 PM
jonrandy.bsky.social
"Who cares if you can’t ship it to the masses, who cares if it’s useless. Make it, create something from nothing, just because you can."

I don't think the curious developer is gone, but we're (sadly) certainly a dying breed.

#programming #development

dayvster.com/blog/dev-cul...
Dev Culture Is Dying The Curious Developer Is Gone
From tinkerers to metric seekers: How the shift in developer culture is impacting innovation and creativity.
dayvster.com
September 20, 2025 at 10:07 AM
jonrandy.bsky.social
"...the business incentives driving consumer AI development remain fundamentally misaligned with reducing hallucinations. Until these incentives change, hallucinations will persist."

theconversation.com/why-openais-...

#ai #openai #chatgpt #llm
Why OpenAI’s solution to AI hallucinations would kill ChatGPT tomorrow
The cure is likely to be worse than the disease.
theconversation.com
September 13, 2025 at 5:05 AM
jonrandy.bsky.social
"More troubling still, it may promote a new procedural ideology of conversation — one where talk is reduced to sounding engaged yet remains uncreative; a functional but impoverished tool of cooperation."

#ai #chatgpt

theconversation.com/chatgpt-only...
ChatGPT only talks in clichés – here’s why that’s a threat to human creativity
AI doesn’t worry about sounding boring.
theconversation.com
September 2, 2025 at 2:26 AM
jonrandy.bsky.social
How OpenAI's ChatGPT Guided a Teen to His Death

youtu.be/tiZ3FyuaHlI...

#ChatGPT #AI
How OpenAI's ChatGPT Guided a Teen to His Death
youtu.be
August 28, 2025 at 6:29 AM
jonrandy.bsky.social
#Superman was really... not good 👎
August 20, 2025 at 4:17 PM
jonrandy.bsky.social
Is it just me, or was episode 2 of #AlienEarth a huge let down visually after the top notch first episode? Blew the budget on the pilot?
August 19, 2025 at 12:17 PM
jonrandy.bsky.social
"Rather than showing the capability for generalized logical inference, these chain-of-thought models are "a sophisticated form of structured pattern matching" that "degrades significantly" when pushed even slightly outside of its training distribution"

arstechnica.com/ai/2025/08/r...

#llm #chatgpt
LLMs’ “simulated reasoning” abilities are a “brittle mirage,” researchers find - Ars Technica
Chain-of-thought AI “degrades significantly” when asked to generalize beyond training.
arstechnica.com
August 12, 2025 at 4:18 AM
jonrandy.bsky.social
Who are all these people "using AI for search"? AI doesn't do search. It assumes you're asking a question and gives a (possibly wrong) answer.

Asking questions, and searching the internet are two entirely different things.

#ai #chatgpt #llm
July 30, 2025 at 2:13 AM
jonrandy.bsky.social
A new study just upended AI safety

Evil is contagious

www.theverge.com/ai-artificia...

#ai #llm
A new study just upended AI safety | The Verge
Evil is contagious.
www.theverge.com
July 29, 2025 at 12:51 PM
jonrandy.bsky.social
Nothing is made to be loved. Only to be bought.

english.elpais.com/culture/2025...
The bewildering phenomenon of declining quality | Culture | EL PAÍS English
Airplane seats are getting smaller and smaller, clothes are unrecognizable after the second wash, and machines now answer our calls. Quality and care for craftsmanship seem to be things of the past
english.elpais.com
July 21, 2025 at 12:31 PM
jonrandy.bsky.social
LLMs are a HACK. The weight of evidence shows they are not thinking, and cannot reason.

#llm #ai #chatgpt

youtu.be/7-UzV9AZKeU...
Oxford's AI Chair: LLMs are a HACK
youtu.be
July 15, 2025 at 4:55 AM
jonrandy.bsky.social
The most interesting things to build are almost always those with little to no practical use. Most other development is repetitive and/or boring.

#webdev #webdevelopment #programming #development #javascript #truths #unpopularopinion
July 9, 2025 at 4:51 AM
jonrandy.bsky.social
So glad I've never in 30 years worked like this... sounds awful, soul destroying.

#development #scrum #sprint #programming #webdevelopment #jira

thecynical.dev/posts/ticket...
Ticket-Driven Development: The Fastest Way to Go Nowhere | The Cynical Developer
When every dev is just doing the next ticket, who’s steering the ship?
thecynical.dev
June 27, 2025 at 8:01 AM
jonrandy.bsky.social
"...in a technological world, information is a problem, not a solution."

- Neil Postman
June 25, 2025 at 4:34 PM
jonrandy.bsky.social
"In a unique experiment, an engineer pitted the latest ChatGPT 4o model against his Atari 2600's chess engine on the beginner difficulty level, and ChatGPT got handily defeated, eventually conceding."

www.extremetech.com/computing/ch...

#ai #chatgpt
ChatGPT Lost a Chess Game to an Atari 2600 | Extremetech
And on the 'Beginner' difficulty level, too.
www.extremetech.com
June 12, 2025 at 1:09 PM
jonrandy.bsky.social
"...they're super expensive pattern matchers that break as soon as we step outside their training distribution"

garymarcus.substack.com/p/a-knockout...

#ai #llm #claudeai #deepseek #chatgpt
June 9, 2025 at 10:42 AM
jonrandy.bsky.social
"...large language models take advantage of the brain’s tendency to associate language with thinking"

www.theatlantic.com/culture/arch...

#ai #llm #chatgpt
June 9, 2025 at 9:32 AM
jonrandy.bsky.social
"The real threat AI poses to education isn’t that it encourages cheating. It’s that it discourages learning."

#ai #chatgpt #llm

www.newcartographies.com/p/the-myth-o...
The Myth of Automated Learning - by Nicholas Carr
AI's real threat to education.
www.newcartographies.com
June 2, 2025 at 8:17 AM