I agree, it’s absolutely absurd, if they’d at least share what specific thing set the alarm bells off i could re-attempt without inadvertently committing a thought crime.
And it’s not like those fable tokens are being refunded, yes the new ones post downgrade are opus rate but the token spend to find out you’re a criminal were fable bucks.
Fable pre-ban wouldn’t continue an old conversation about brain chemistry, sooo I guess they’re dusting off old tricks now that their new robot is offline.
“Wow, are you saying I kind of singlehandedly built the kind of stack they use at Google? If engineering departments only knew… how can I get some CTO to hire me as a chief engineer?”
was probably when chatgpt should have said - no you built what seems like an interesting/capable php framework.
Your homepage looks cool but doesn’t explain what the product does.
This is just one man’s take:
“The Human-AI Operating System” is marketing vapor. “Harmonize your workflow” tells me nothing. The three feature boxes—“Intelligent Automation,” “Seamless Integration,” “Adaptive Learning”—could describe any SaaS product from the last five years.
I had to scroll to the demo video to understand you’re building event-driven workflow automation. That should be the first sentence.
Compare to your explanation: “Shopify order comes in, system checks integrations, shows you the route, one button completes all actions.” That’s clear. That’s what someone needs to know in five seconds.
Your homepage buries the actual value under abstraction. No one knows what “harmonize” means operationally. They don’t know what problem you’re solving or what actions the system takes.
Strip out the philosophy. Lead with the mechanism: “When business events happen, we route them across your tools automatically. One click fulfills orders, notifies teams, updates calendars. The system learns your patterns and gets smarter over time.”
Then show the integrations—Gmail, Shopify, Stripe, Slack, Asana. People need to see “oh, this connects the tools I already use.”
The cosmic language (“reshape how humans and AI collaborate”) actively obscures what you built. You’re competing for attention against a thousand AI startups using identical words. Specificity is your only advantage.
What does the product actually do in the first 60 seconds after someone signs up?
I’ve run into a similar issue, compacting is pretty much worthless and leads to a lot of churn.
I have started having Claude pay constant attention to usage and assess whether the next batch of tasks flirts with compaction.
If we’re anywhere close I have Claude create a next steps doc with meticulous notes along with a prompt that I can give to the next llm to hit the ground running.
Then kill the session and start fresh.
At least this way I know there’s a fighting chance of not falling into a death spiral of the llm guessing about what it’s supposed to be doing as if it hasn’t just had a self inflicted stroke.
I work for Amazon (AWS for 4 years then “the website” side of the house for the last 3)
The previous commenter is correct, there is no NOC or devops team and I’ve not encountered a Devops job family and I’ve never seen terraform internally.
Within AWS, the service teams that work these outages are the same ones that design the service, fix bugs, deploy the pipelines, be oncall, etc. the roles that fill these teams are pretty much one of three types: nde, sde, sysde. They typically use cdk if they’re doing AWS things, else they’ll use internal tooling.
The job you posted is a customer facing consultant like role - customers use terraform so having a customer facing consultant type that knows how customer-y things work is a good decision.
100% I drive a shell all day long, there’s no reason to drop to a browser to copy and paste.
Just create new folder, init, start the conversation and have it take notes/write docs. It’s a splendid way to work.
Plus if you name the folder something meaningful you can always go back without needing to scroll through vaguely named chat conversations
Correspondence, keeping records and making appointments are attention destroying activities.
For a few year stretch I worked with a team project manager who wasn't technical at all. She'd setup the meetings that I needed to have, diligently take notes in those meetings and then we'd get together to recap everything figure out dates and all that. Then I'd get back to nerding out but with a ordered punch list that I knew Donna needed me to get done.
It was the best.
But the pmo wanted to do things a different way so their engagement model moved from embedded PMs focused on team deliverables to individual project focus.
Now instead of Donna there to assist with correspondence, keep records, make appointments, and carry out similar tasks I had an army of fragmented Donnas each focused on their one thing and each needing oversight to understand my involvement and external dependencies.
The move probably gave the pmo a more modern feel and better reporting but it killed productivity for ADDs like me since a large portion of my time each day became correspondence, keeping records and making appointments instead of cranking out nerd units.