As is already revealed with Meta leadership knowing that they make 7billion a year on scam ads. They even calculated that global regulations and fines might cost them 1 billion.
So fines and regulations are priced in as a fraction of the net earnings.
> Never mind the fact that AIs of the LLM-variety haven't and aren't going to find solutions to mathematical problems.
This is empirically wrong as of early 2026.
Since Christmas 2025, 15 Erdos problems have been moved from "open" to "solved" on erdosproblems.com, 11 of them crediting AI models. Problems #397, #728, and #729 were solved by GPT-5.2 Pro generating original arguments (not literature lookups), formalized in Lean, and verified by Terence Tao himself. Problem #1026 was solved more or less autonomously by Harmonic's Aristotle model in Lean.
At IMO 2025, three separate systems (Gemini Deep Think, an OpenAI system, and Aristotle) independently achieved gold-medal performance, solving 5 of 6 problems.
DeepSeek-Prover-V2 hits 88.9% on MiniF2F-test. Top models solve 40% of postdoc-level problems on FrontierMath, up from 2%.
Tao's own assessment as of March 2026: AI is "ready for primetime" in math and theoretical physics because it "saves more time than it wastes."
You can disagree about where this is heading, but "haven't and aren't going to" doesn't survive contact with the data.
Disclosure: I'm on the Mautic leadership team — I lead the marketing team and recently joined the Council (our governance board).
Happy to answer any questions about the situation. A few things worth noting:
Our finances are fully transparent: monthly reports are public via Open Collective and our Open Startup Reports. The shortfall came from a small number of key members re-budgeting, not from declining usage. Mautic's install base is actually growing. We've already cut costs significantly (40% reduction in Project Lead hours, cancelled a developer hire, suspended events). The $140K essential budget covers codebase management, infrastructure, and trademark protection.
The broader question of how open-source projects that serve as critical infrastructure can sustain themselves financially is something I think about a lot. We're trying to be transparent about navigating it in real time.
Disclosure: I'm on the Mautic leadership team — I lead the marketing team and recently joined the Council (our governance board).
Happy to answer any questions about the situation. A few things worth noting:
- Our finances are fully transparent: monthly reports are public via Open Collective and our Open Startup Reports
- The shortfall came from a small number of key members re-budgeting, not from declining usage — Mautic's install base is actually growing
- We've already cut costs significantly (40% reduction in Project Lead hours, cancelled a developer hire, suspended events)
- The $140K essential budget covers codebase management, infrastructure, and trademark protection
The broader question of how open-source projects that serve as critical infrastructure can sustain themselves financially is something I think about a lot. We're trying to be transparent about navigating it in real time.
“Kagi Search is an ad-free search engine that will actively down-rank sites with lots of ads and trackers in the results and promote sites with little or no advertising”
There are so many config options. Most I still need to truly deeply understand.
But this one isn't? I'd call myself a professional. I use with tons of files across a wide range of projects and types of work.
To me file paths were an important aspect of understanding context of the work and of the context CC was gaining.
Now? It feels like running on a foggy street, never sure when the corner will come and I'll hit a fence or house.
Why not introduce a toggle? I'd happily add that to my alisases.
Edit: I forgot. I don't need better subagent output. Or even less output whrn watching thinking traces. I am happy to have full verbosity. There are cases where it's an important aspect.
You want verbose mode for this -- we evolved it to do exactly what you're asking for: verbose file reads, without seeing thinking traces, hook output, or (after tomorrow's release) full subagent output.
Sorry to rain on your parade. I wanted the original verbose mode for those moments I needed a truly verbose output. And I wanted to know, at a minimal glance, what files are being read and put into context in nearly any other situation.
I exactly do not need a "verbose" mode, that lost all value to me as a replacement for something it still is no good at replacing.
You actually argue, that I do not loose anything, when in fact your product just got made worse in two significant areas. And you keep arguing, that shooting the product into one foot is solved by shooting the other foot. Sorry. Not working for me.
Will be evaluating your competition. Was on the cusp of upgrading max to the higher tier. Now? No chance of that happening.
Quite a few - and I know I am only speaking for myself - live on my different computers. I created a few CLI tools that make my life and that of my agent smoother sailing for information retrieval. I created, inspired by a blog post, a digital personal assistant, that really enables me to better juggle different work contexts as well as different projects within these work contexts.
I created a platform for a virtual pub quiz for my team at my day job, built multiple pandingpages for events, debugged dark table to recognize my new camera (it was to new to be included in the camera.xml file, but the specs were known). I debugged quite a few parts of a legacy shitshow of an application, did a lot of infrastructure optimization and I also created a massive ton of content as a centaur in dialog with the help of Claude Code.
But I don't do "Show HN" posts. And I don't advertise my builds - because other than those named, most are one off things, that I throw away after this one problem was solved.
To me code became way more ephemeral.
But YMMV - and that is a good thing. I also believe that way less people than the hype bubble implies are actually really into hard core usage like Pete Steinberger or Armin Ronacher and the likes.
> Quite a few - and I know I am only speaking for myself - live on my different computers
I use AI/agents in quite similar ways, and even rekindled multiple personal projects that had stalled. However, to borrow OPs parlance, these are not "houses" - more like sheds and tree-houses. They are fun and useful, but not moving the needle on housing stock supply, so to speak.
I didn't even put in my last hotel bill, because my time was worth more to me than navigating a shitty, always borderline broken excel file that once filled out leads to a back and forth of clarifying questions always with an undercurrent of "them" trying to "teach me how to be a good form fill monkey".
If I should win the next pitch, I will definitely have on e of my team work on setting up a shadow system, a web application that produces pdfs for the finance department that are indistinguishable from the XLS based ones. But with a better CX for me and my people.
LLMs could be a good option to navigating this sludge. Fight fire with fire.
Another option is work smarter (not harder, because nothing I do is anywhere near “hard work”) to get into a position where you can tell them to get fucked. Don’t want to pay my hotel bill? Oh well, good luck finding someone else to rework your auth system. Call me back when the outsourced monkeys you hire end up putting you in the news for a security breach. But at least you saved a few hundred bucks on hotel fees, great job!
This is something management and executive positions do on a continuous basis - using their position and “prestige” to commend respect and bend the rules. But as an engineer with context of a critical system you often have more leverage, it’s just a matter of using it strategically (as engineers we initially start out playing the good game, but the thing is that everyone else is trying to fuck you - the challenge is learning to fuck back).
I keep a beginner’s Python book in reserve for those conflictual meetings where some idiot beancounter or manager has a problem with me. When I’m ready to walk (and at this point I have a very short fuse for obvious bad faith), I offer it to them as a tool to help them finish my job; not a single soul has yet to take me up on that offer. Some idiots suggested me the way to the door a few months later (offer gladly accepted, and replacement gig acquired) and watching from a distance it’s clear they would’ve been better off actually taking that book off me - either for themselves or the idiots they tried to hand my tasks to.
The only way to enact change is to actually make the noxious behaviour costly. If you take on the costs yourself there’s no reason for them not to persevere with their misguided strategy.
I can mightily relate. The would pay my bill. I just value my time higher than 190€ for navigating a shitty XLS file and the unnecessary comms afterwards. Yes, abso-f*ing-ly a position from privilege. That's why at the next opportunity, me having to travel in 2 weeks, I will use the time and my trusted LLM accomplice to engineer a solution that afterwards will probably be used by a few more people.
Why do you assume it is finance departments making these decisions?
Finance departments get hit with the same clueless 'cost saving' as everyone else! Often because the savings are so 'obvious' to MBA wielding middle managers, the decisions are made without Finance advice.
It is also often just a corporate politics thing. Your department has to save money, but there is no measure of how much money you cost other departments. In my experience this is more often done to Finance than by them. Pushing more work from frontline departments to back office departments sounds great to an Ops manager, and this frontline departments often contain people who are good at selling their idea.
I don't assume. I spoke/wrote form experience over many companies big and small (as an employee as well as a consultant). I have seen my fair share of crap. I am an just an old and grumpy guy.
So fines and regulations are priced in as a fraction of the net earnings.
https://mashable.com/article/meta-7-billion-dollars-scam-ads
reply