It depends. For drivers, it’s basically been captured.
For say, peer-to-peer car or vacation rental (e.g. Turo, Airbnb) you have some one man companies but the top hosts on the platform is usually an established company often with a staff of employees.
It’s not until you tap through to see they have tens of thousands of trips, dozens of cars in each city, etc.
So when you say Uber driver - it may work in the “driver” category only. In pretty much the rest of the gig economy there is a lot of room for both solopreneuers and SMB.
The “gig economy” may well end up just being “the economy”.
I don't really get the idea of needing converts when nixpkgs is basically the largest and most up to date.. Using another Linux as a system engineer you would certainly know how to build outside the package system because you would have to.
Interesting theory and I can see how you get there.
But if drainage changes were the primary cause, we would've seen drought severity increasing across the time period where straightening/etc were happening... instead the drought severity increases begin _after_ that period, better aligned with temperature changes.
Additionally, your analogy to a water leak from a tank doesn't hold, because the water recharges over time...
Further, your complaints about immigrants also doesn't make sense, as water usage has been going down (and residential utilization is quite low, altogether).
As such, my theory is that you're just a right-wing crank who wants to find a reason to deny global warming, and to hate immigrants.
It would be frustrating, but the group that is excluded from online services today tends to be low status and may benefit greatly when a tech elite with connections, status and special interest groups finds itself excluded.
I was once tracking habits with an excel file a while ago. Works well too! An App can be move involved. For instance, excel files or papers can not send you reminders...
The ephemeral-pod-plus-restore-on-init pattern works right up until it doesn't, and the failure is quiet, so "no issues so far" is unfortunately consistent with both a correct setup and one that is going to bite.
Litestream assumes a single writer. It replicates one SQLite file's WAL to object storage; it is not multi-master. If you ever end up with two pods writing, you get two databases diverging and replicating over the top of each other, and nothing errors. You find out when you notice writes going missing.
The two ways people get there without meaning to:
- replicas gets bumped above 1, sometimes by an HPA nobody remembered was on.
- RollingUpdate. It briefly runs old and new pods together, which is enough. Recreate avoids it at the cost of a restart gap.
The other thing worth knowing is that replication is asynchronous with a default sync interval of one second, so a hard pod kill can lose whatever was written since the last sync. Fine for a lot of workloads. Not fine if the thing has quietly become a durable transaction log for something that matters.
LiteFS was the answer for genuinely multi-node, though I would check its current maintenance status before building on it.
first try was hidden? OpenAI sent me a “cyber abuse” warning. No OpenAI-directed chat could have caused it. None. The only relevant activity was Codex tasks on my private local model, deliberately routed to my machine, not OpenAI.
I call it data exfiltration because ordinary use fits: data crossed the private provider boundary I selected and went to OpenAI without my informed authorization. An undocumented automatic background feature does not authorize it.
Codex Memories searches old chats without restricting them to their original provider, then processes the selected chat with the new session’s provider. A local-provider chat selected in an OpenAI session is uploaded as a new prompt.
Selector requests eligible old chats with model_providers: None. Worker uses the triggering session’s current provider. It serializes the chat and sends it there as a new model request. I tested stock Windows Codex 0.150.0-alpha.12.2 using a synthetic chat with unique canaries, not more real private material. The source chat was labeled non-OpenAI; analytics disabled; every OpenTelemetry exporter none. No setting keeps Memories while blocking this. The only kill switch is [features] memories = false. Codex still sent a 38,095-byte WebSocket response.create frame to chatgpt.com/backend-api/codex/responses, marked request_kind = "memory". Complete payload: environment-context message (working directory, workspace, shell, date, timezone, filesystem policy); user message; tool-call name/arguments; tool output/error including runtime paths and policy text; assistant reply. That is the chat, tool activity, and local-machine information, not fucking “telemetry.” OpenAI returned a completed gpt-5.6-luna response. Its memory repeated the unique user, assistant, tool, private-text, and synthetic-email canaries, unarguably confirming OpenAI received and processed content they never should have touched.
That explains the warning: OpenAI-directed chats had nothing relevant; private local-provider chats did; OpenAI issued it. Memories was the hidden channel policing private local-model activity I deliberately routed away from OpenAI.
OpenAI has no right to take chats from a provider I chose for privacy, later upload/process them in a background job, and treat them as account activity. A local provider is a data/consent boundary; crossing it behind the user’s back defeats its purpose.
This is like opening someone’s bedroom door without permission, disliking what you see, then punishing them. You had no right to open it. Any strike, warning, restriction, or other account action based on it is fruit of the poisonous tree.
If intentional, the legality and adequacy of disclosure are debatable. OpenAI should identify the exact consent authorizing upload, where crossover was disclosed, what it retained/derived, and exactly how private chat content was used for the warning.
User-facing memory docs discuss local memory storage/controls; the developer README says filtered chats are sent to “a model.” Neither warns that one provider’s chat can later be uploaded to another, including OpenAI.
The fix is simple: bind memory processing to the source chat’s provider; if unavailable, skip it. Never send a local-provider chat to OpenAI without explicit, destination-specific consent. Log every transfer locally so users can see where each chat went.
OpenAI owes users an official public response: disclose the scope of this exfiltration, backend use, whether content was used for account moderation, and how many warnings, strikes, restrictions, or other negative account actions depended on it.
Every affected user must be addressed: reverse those actions; notify users directly; explain which private-provider content was used; delete uploaded/derived records. Quietly fixing the selector is not enough after penalizing users for content OpenAI had no right to see.
Universities are the type of place you would automatically expect to teach the fundamentals. They do it in every other subject. You should have "coding like a caveman" back to back with "coding like a spaceman". You can make the hand coded class have two 1.5hr in-person labs each week. With an AI class along side of it. Don't wait on either skill, both need to start freshman year.
First year AI is teach assistive coding using llms with open source models. Setup a workbench llm for each student to have with an openweight model on it. Teach different ways of getting good results and go over setting up the whole environment yourself.
The next year do a 4hr lab in a lab room with office chairs, variable height desks, and model it off an open office. No special dress code, but name tags and people will talk to each other with the openness of dialog at a good workplace. This even gives the TA usable experience. You could introduce different structures with leads or co-equals or any useful team structure.
People need to be taught in-person working skills because companies don't want to hire people without experience. This also helps someone be able to interview for a work from home job.
Finally someone mentions corporate influence and enrichment of three future trillionaires.
AI is an inherently fascist and thought controlling technology and we can see that most of Debian has already been brainwashed.
In the beginning, Debian fought against Microsoft. The good and intelligent people left and Debian is just another arm of big tech. It is nauseating to read the low intellectual level of the conversations and the feigned naivete of the corporate shills. Debian is the new Microsoft now.
So what remains? Gentoo, LinuxFromScratch, FreeBSD, NetBSD!
> I always use sudo to manage the few docker containers I need
I'm afraid that isn't really any better. If the attacker is in position to exploit membership in the docker group, he already has access to the user's .bashrc. He can simply write a function called sudo that wraps the real command and records your password[1]. Unless the user always invokes sudo with /usr/bin/sudo, docker group membership doesn't really make a difference.
Once malware runs as an administrator, getting access to root isn't really that complicated. The boundary between wheel and root is more or less security theater.
Edit: Oh well, I see now that others have made the same point (https://news.ycombinator.com/item?id=49500588). With the same wording even. I'm half-way tempted to delete my comment so as not to look like a plagiarist, but it seems many posters are unaware of the vulnerability, so I'd leave it here.
I understood your original comment just fine, I just thought that the point you were making was stupid.
By "trend nowadays" I assume you're just choosing not to count the 700k layoffs in tech sector since 2022?
Show me all of the new products Google, Meta, Amazon, Microsoft etc have "spawned" since force feeding AI to their remaining staff. 100 different permutations for sycophantic AI models and fact simulators don't count.
Has your dream of being in band been unfairly denied by a lack of talent, discipline, or the ability to play any instrument whatsoever? Hope has arrived.
I am accepting prompt-to-audio auditions for my cartoon band.
How it works:
- I fine-tuned an audio model based on my own musical work and tastes
- you type prompts that create riffs, not entire songs, within that palette
- I arrange the riffs and write songs around them
- I use my human voice to sing over them
- people rant about AI destroying human creativity anyway
So far the primary reactions have been either 1) they hate this and think it shouldn't exist, or 2) they want to buy merchandise. Both are good signs.
Is the future now that we get rambling report summaries talking about agents, graders and so forth without ever describing how they are set up? A human launches all this.
And then the original reports linked to are hidden on the now unreachable x.com. And they don't have a problem with that.
I hope it attracts attention. Everything bad that rational people predicted about the DHS when it was created has come true, but people seem to have become comfortable with their civil liberties taken away.
For say, peer-to-peer car or vacation rental (e.g. Turo, Airbnb) you have some one man companies but the top hosts on the platform is usually an established company often with a staff of employees.
It’s not until you tap through to see they have tens of thousands of trips, dozens of cars in each city, etc.
So when you say Uber driver - it may work in the “driver” category only. In pretty much the rest of the gig economy there is a lot of room for both solopreneuers and SMB.
The “gig economy” may well end up just being “the economy”.