Want to wade into the rainbow-ridden surf of the abyss? Have a sneer percolating in your system but not enough time/energy to make a whole post about it? Go forth and be mid.
Welcome to the Stubsack, your first port of call for learning fresh Awful you’ll near-instantly regret.
Any awful.systems sub may be subsneered in this subthread, techtakes or no.
If your sneer seems higher quality than you thought, feel free to cut’n’paste it into its own post — there’s no quota for posting and the bar really isn’t that high.
The post Xitter web has spawned so many “esoteric” right wing freaks, but there’s no appropriate sneer-space for them. I’m talking redscare-ish, reality challenged “culture critics” who write about everything but understand nothing. I’m talking about reply-guys who make the same 6 tweets about the same 3 subjects. They’re inescapable at this point, yet I don’t see them mocked (as much as they should be)
Like, there was one dude a while back who insisted that women couldn’t be surgeons because they didn’t believe in the moon or in stars? I think each and every one of these guys is uniquely fucked up and if I can’t escape them, I would love to sneer at them.
(Credit and/or blame to David Gerard for starting this.)
Saw a guy watering his lawn this morning. Just standing there, hose in hand, dumping potable water onto grass that exists for no reason other than to be looked at and complained about.
Sir. Do you understand that a single hyperscale data center can drink millions of gallons a year keeping GPUs from cooking themselves while they generate a poem about a sad robot? That water has a HIGHER calling. That water could be evaporating off a cooling tower in service of someone’s RAG pipeline that returns the wrong answer with tremendous confidence.
And here you are. Hydrating Kentucky bluegrass. In a region where the grass was never supposed to grow in the first place.
I asked him if his lawn had an SLA. He said no. I asked what his lawn’s uptime commitment was. He looked at me like I was the unreasonable one. Meanwhile that turf is sitting at four nines of being green and producing exactly zero tokens per second.
We are pouring concrete across three states to host inference workloads, and this man is allocating municipal water to a crabgrass cluster with no monetization strategy. No usage-based billing. Not even a freemium tier.
Anyway I reported him to nobody, because there’s no one to report him to, which is honestly the most damning part of this entire ecosystem.
Touch grass, they said. He did. Look where it got us.
NOT EVEN A FREEMIUM TIER. that got me.
this man is allocating municipal water to a crabgrass cluster with no monetization strategy
This is poetry, AI could never
Old: AI Winter
Bold: AI Ragnarok
Botterdamerung
Noice
AI Winter but with the same connotations as a nuclear one
The new Claude model will silently decide whether what you asked it to do is in line with anthropic ToS and silently corrupt your prompt if it doesn’t like what you’re asking. It’s couched as a “safety countermeasure” but it is presumably to stop Chinese labs trying to scrape synthetic data.
We’ve moved from ‘accidental’ hallucinations to deliberate misinformation and you’re paying $$$ for the privelige.
Claude can now be silently nerfed. Anthropic has decided it won’t tell users when this happens.
considering how many habitual llm users can’t tell good from bad output anyway, they always could have done that
To what extent should one trust a statement that a program is free of Trojan horses? Perhaps it is more important to trust the people who wrote the software.
Ken Thompson, Reflections on Trusting Trust
I know this outcome was inevitable after software became a mass market thing, but it’s still rather depressing.
“You’re holding it wrong” as official policy
deleted by creator
Wake up babe new genre definition just dropped
https://stoates.substack.com/p/programmer-science-fiction
Includes all the usual suspects, but notably doesn’t mention Ken MacLeod whose Fall Revolution series is probably too socialist[1]. Also avoids discussing Stross post-Singularity Sky.
[1] MacLeod’s Corporation Wars trilogy has immersive VR, artificial conciousness rebelling against authority, and literal p-zombies but is also very anti-fascist, so no wonder it’s not mentioned (also, it’s unfortunately not very good)
Viral: “read a second book”
Spiral: “read a second Borges story”
The author of “Death and the Compass” and “Emma Zunz” is unrecognizable from the description there.
@gerikson @techtakes Technical nit-pick: “American hard science fiction space opera like Timelike Infinity is also influential”— Timelike Infinity was written by Steven Baxter who is *very* English indeed. Best contextualized as mid-period Interzone generation hitting its imperial phase.
This is a new method of bootlicking: if you as a FLOSS developer don’t use LLMs to fix vulnerabilites identified by LLMs, you’re being unethical
Lol what a weak argument. Not only does this limit your llm use to only bugfixing (which is what the op is limiting their use to right?), it also ignores how a few big recent outages were prob caused by llms. And it treats ethical concerns like some sort of numbers game. We have one ethical concern for and one against so it cancels out.
And it leads to ‘Mengeles experiments were not unethical because some of the torture he did actually provided valuable insights on the extremes a human body can go through’
“It is your moral duty to send as much money as you can to Sam Altman”
-
"The instructions say “You are an agent that writes the various article sections for an encyclopedia entry on Joseph Kanuku.”
-
“[Note: Britannica is encyclopedia, but instructions say no Britannica. Wait, replace with another.]”
-
“Wait, instructions say NEVER cite Wikipedia. Oops. Let me adjust.”
-
“The instructions say prioritize peer reviewed, books, etc.”
-
“Wait, but instructions say never cite social media, so omit that last part. Wait, adjust. Since Instagram is social media, omit that.”
-
“Wait, instructions say avoid “References” as a section, but for completeness, I’ve omitted it from structure.)”
-
“The guidelines say “Include any of the following where relevant: - Factual details…” under Missing Information or Knowledge Gaps Examples. And in format, it’s to list only critical issues or all missing info or knowledge gaps.”
-
I guess a certain country is using LLMs to try to engage people 1:1 to change their opinions of said country.
Could be a LLM, could be a call-center worker in Kenya or Mumbai or Manilla. Its an old move.
Good point. Now I imagine someone trying to pass the turing test by having their bot pretend to be a sales person.
Anthropic insists that you should use their demon-as-a-service. Hackernews debates the finer points of pentangle construction.
https://news.ycombinator.com/item?id=48463808
Edit: a hackernews is tired of simonw:
Today we’re launching Claude Fable 5: a Mythos-class model that we’ve made safe for general use.
oh fuck off. going full SCP in their press releases now. insufferable.
Babe wake up, new model that’s too dangerous for public consumption but not really just dropped.
On a related note Mythos has been released and as pointed out, the original “GPT-2 is too dangerous to be released” post was written by Dario Amodei, Daniela Amodei and Jack Clark, who are all Anthropic’s founders. Same ploy, different model
I’ve also seen mention of “Fable” and I’ve been too sick to go check up on whatever the fuck that is
Mythos with extra safety dust sprinkled on top and they listened to The Safety Dance while writing the system prompt. Also a hand painted sign on the front that says “no chinese were serius”
Hope you get to feeling better fast!
after calling the con about a super hacker AI the “mythos”, now they’re calling the next spiel “fable”? are they like, just taking the piss at this point?
Claude v6.0, “Tall Tale”
Claude v6.1 “Big Fish Story”
They want to invent the AI Named For Untruth from the hit Star Trek: The Next Generation episode “Do Not Invent the AI Named for Untruth”.
just taking the piss at this point?
always-have-been.jpg
okay damn you o7 I checked
Claude Fable 5 introduces our 5th model generation for your most ambitious work. Tackle days-long, complex, and asynchronous tasks previous models couldn’t sustain.
folks, can I get a “please god pay us for even more tokens” for 500?
and they listened to The Safety Dance
not sure they’re cool enough even to do that
Hope you get to feeling better fast!
ty! rest is helping, but why are sinuses
remarkably educational, nice
It’s pretty good! The kids want to be influenza for dragoncon lol
It’s on netflix in the states
Let’s all just agree that it’s a beloved RPG game by Lionhead Studios and we’ll leave it at that
lol
Time is a dumb circle
Is AI profitable yet? a website that tracks the spending and revenue of leading AI companies. The answer is so helpfully provided at the top of the page
It’s also fun to sort by pnl and see just how far the gap is for others that aren’t nvidia
New pride month Odium Symposium episode. We go through the life and activism of notorious homophobe and orange juice queen, Anita Bryant
I’d been betting on one of the promptfarmer companies being the ones to set off the bubble-popping chain reaction, but looks like musk is so desperate for cash that his coterie of grifts are speedrunning for the underdog position
3 weeks after the “as soon as Friday” news, OpenAI has followed Anthropic and confidentiality filed their draft S-1. But they sure don’t sound confident about it. Whole post in full (sans legal fine print):
We recently submitted a confidential S-1. We expect it to leak so we’re just announcing it. We have not decided on timing yet; it may be a while because there are things we want to do that are likely easier as a private company. But it’s a complicated set of tradeoffs and this gives us the option to go public sooner if that ends up being best.
May, likely, if… Those are some weight-bearing subjunctive clauses.
Edit: also Altman’s eyeball tracker company is doing layoffs now
Don’t they have their CFO not even reporting directly to be CEO? I would bet that there’s a ton of internal dissent about timing and strategy of how to cash out.
That CFO thing was definitely the case, at least a few months ago. I’m sure you’re right about the internal chaos, even if that CFO drama has changed, and it would align with how non-committal this announcement is.
I would love to be a fly on that wall.
Wait why would they expect it to leak? (assuming that mentioning it isn’t just some marketing stunt, which it probably is)
They asked a chatbot and the chatbot is trained on shadowrun and cyberpunk2077 sourcebooks.
Ed Zitron might have got hold of it, if his bsky is anything to go by.
We could be feasting soon!
IANAL, but my understanding is that companies are allowed to keep confidential the fact that they even filed the S-1. That’s how I read OpenAI’s statement. But it’s not completely clear what “it” means in “we expect it to leak”.
Draft filings involve feedback from the SEC, so I think they may be throwing shade on government employees, whom they can’t fire or control directly.
But I’m also thinking they may be salty about the aforementioned “as soon as Friday” articles, and then Anthropic beating them to filing.
Hard to tell how much of it is what, they’re toxic inside and out.
Molly White continuing to slam it out of the park with a pivot to AI. As always, worth a read in full, but this intro bit stuck out to me after reading lots of inane blather on how crypto and AI are different:
Continuing to track only crypto would mean missing half the story. The same operatives are running both campaigns. Josh Vlasto, longtime adviser and spokesperson for Fairshake — the cryptocurrency super PAC network responsible for the bulk of crypto’s 2024 spending — is now simultaneously heading Leading the Future, a pro-AI super PAC network. Chris Lehane, the political consultant and Coinbase board member who helped establish Fairshake and famously told Coinbase employees who questioned whether a crypto voter bloc existed that they would simply invent one, is now also an OpenAI executive and one of the people behind the Leading the Future PAC network. The same venture capital firms are funding both: Andreessen Horowitz, a crypto heavyweight in the 2024 elections, is now splitting its political spending across crypto and AI PACs.
Good timing since it seems even Crypto folks are getting out of Crypto.
Literally the only pivot to AI that I am happy about. Molly is GOAT
Coworker got fired because he used AI to plan for a site installment of our product. The AI made a very nice looking plan but it failed to include enough packing material so nearly half of the units arrived broken. Boss still thinks AI is going to revolutionize work for the better though.
Came across this today: the purpose of a system is what it does. Based on that I would say the purpose of Claude was to make mistakes and get my coworker fired.
to be fair, the plan would have worked if you had included the allotted number of goblins in the box in addition to the packing material.
