π Newsletter #010 - LLMs and: Profitability, Security, and not Leveling the Job Application Playing Field; plus Good News on Pushback & Digital Sovereignty
β¨ In This Edition β¨
- π Short-Form Content: AI CV's: (Some of us) are damned if we do, damned if we don't; Is it AI, or is it a security liability!?, Is AI profitable?, and why the AI bubble is not like the internet bubble.
- π° Long-Form Content: The Alternatives Pt. II post on Social Networking has gone BearViral (I think)
- π Good News!: Protest works (a] vs AI data centres, and b] vs Not So Good bills); the EU Commission is bullish on Open Source; more politicians and comedians are speaking out against AI.
- π― The Post-Script: Reading Shawn Smucker's essay titled "Please Use AI."
Mashup of two images by Markus Winkler & Tyler Prahm on Unsplash, cropped & glitched.
π Short-Form Content
Chatbots ("AI")
- Not all AI-generated CV's are created the same:
"Reviewers of Emilyβs rΓ©sumΓ© were 22% more likely to question whether the individual could be trusted compared to James. The female candidateβs CV was also twice as likely to raise doubts about her competence and ability to do her job."
Amazing! At every turn we get told "You must be open to using LLMs if you want to not get left behind. Rocket emoji! ... well, no, not like that." Damned if you do, damned if you don't. Some of us, anyway.
- All major AI models violate EU regulations. Surpriiiiiise!
- The, uh, "best" of them all was Claude Opus 4.7, which was in compliance slightly more often than if you tossed a coin repeatedly!
- These AI models chatbots are so helpful, not to mention secure, that they give you access to instagram profiles you would like to access if you ask nicely enough.
- ... not, like, your own profile that you lost access to. No, no. Any profile! Including those that aren't your own! Back in my day we called that a "significant security flaw," but since it's "AI," this seems to all be fine and dandy.... Unless I am missing something.
- isaiprofitable.com does exactly what it says on the tin, and it's glorious. On that note:
- No, LLMs are not the economic powerhouse that a lot of people (and companies, and CEOs of companies) are trying to make you believe.
- There is no way to break even for the LLM industry. That bubble is going to burst, and tech companies will be hit with the triple whammy of rising token costs, no measurable increase in value, and fewer employees to clean up the mess AI made.
- Hey, did you know that Grok is totally going to be worth the entire economic output of the US? Like, totally. You better get onboard right now.
- Some police in the US are apparently scanning social networks for activity "critical of AI."
- One of my least favourite side effects of the AI boosterism that is happening all around us is that now some people feel emboldened to, instead of answering a question you asked them, plonking the question into an LLM and sending the screenshotted response back to you. Cooooooool.
- I wasn't able to track down the original source, but this reminds me of that meme that goes "if you're going to send me LLM output, please just send me the prompt so I knew what you meant to say."
- I guess we're at a point where "AI radio hosts" exist, after apparently someone somewhere thought this was a good idea. Anyway, it's already gone severely off the rails.
- Another excellent piece by Cory Doctorow: The AI bubble isn't like the internet bubble. For starters, nobody had to force us to use the internet:
What we never saw back then were stories about how managers had to monitor workers to ensure that they were using the web as much as possible. No one had to force workers to find ways to integrate the web into their workflows. In other words, the story of the web at work was the opposite of the story of AI at work. Today, you can't turn around without reading a story about bosses who are threatening to fire workers if they don't increase their AI usage
- And on that note - Another "success story" for when the metric becomes the target:
- Not only are software engineers and coders forced to use this "tool," and are then also pitched against their colleagues in terms of how much they use the tool, but no! We are now apparently at a point where burning down the rainforests using stolen content is incentivized to such a perverse degree that employees feel the need to set up "agents" that burn through their tokens, just because.
Non-Chatbots
- This is just... so predictable: Meta adds Instagram, Facebook, and Whatsapp subscription tiers. Why is it predictable? Enshittification. π© See also #31 of the Let's Talk Tech Thursday newsletter.
π° Long-Form Content
The Alternatives Pt. II post on Social Networking has gotten quite a bit of traction - 120 updoots on Bear is BearViral, right!?
Share it with some folks who aren't all too happy with their current Big Tech social networks!
π Good News!
"I reject the idea that we have to be slaves to surveillance capitalism in order to participate in the modern economy."
-Wab Kinew, Premier of Manitoba. Yes!
- DuckDuckGo is a very capable search engine that has seen a significant increase in "No AI" search traffic after google did that thing where they've decided they'd effectively be killing off search as we know it.
- Apparently religious exemptions re: forced AI usage at work are a thing now!?
- Protesting AI centres works: After hundreds protest Hamilton data centre proposal, committee denies Steelport planning application
- We love seeing the EU Commission be even bigger on open source now, including it in their strategy for technological sovereignty.
- The Canadian government apparently got so much pushback on Bill C-22 that they are open to amending it. Good! (Also, if anyone tries to convince you that we should push this through because "Europe is doing this already:" That is False.)
- Fun, free, and easy ways to add visuals to your blog post/article that don't involve AI.
- While age gating still isn't it (but keeps being proposed nonetheless), we're at least seeing some American states move to exempt open source software (including Linux operating systems) from the requirement.
- "I'm here to tell you the mission of your generation is to destroy AI." - Ronny Chieng, noted Daily Show host, in a speech to students during Harvard's Class Day. He also said the following:
"Your generationβs upcoming battle wonβt be humans against AI; thatβs at least two months away. β¦ Itβs going to be people with substance versus people with shallow knowledge. Itβs going to be mastery versus faking it. Itβs going to be people with good taste versus tacky. I trust you will put in the work necessary to be on the right side of those battles."
π― The Post-Script
Reading Shawn Smucker's beautiful essay titled "Please Use AI." Yes, I know the title sounds out of place in this newsletter. You should read it anyways.
Follow this blog via RSS | Find me on Mastodon | Bubbles
Recent Posts