TrueDuality

I'm super excited for you and this, but this is mostly a highlight list of who is using your library. I was kind of excited to read about the actual changes to the library but I suppose I can go read the changelog on my own.

58% of the time models do this 100% of the time!

Tis a silly question. This has the combination of being something that the populace either doesn't care about at all (so no opposition from _constituents_) or rapidly supports (basically any gamer with some form of nostalgia that has gone unmet). As an added bonus it has the potential to set a legal precedent over similar more serious software such as SaaS offerings with a client component.

All text, but none of these were a one-shot task. There were a few serious gotchas that you need to pay close attention to. It understands problems VERY well when it actually controls the viewport and takes screenshots of the part, unfortunately it REALLY resists that instead preferring to make up a bunch of equations to justify/validate it but these were almost never correct. It frequently got things like axis backwards so when it would test for a potential collision it would test in the wrong direction. Guidance specifically for the axes might help there, I suspect it's assuming different real-world axis (maybe more aligned with the game industry axis than modeling or CAD axis standards).

When I approached those two projects I took different approaches. The workbench I had clear measured physical requirements and desires but no specific design in mind (such as number or size of drawers). I had a long back and forth working out a purely textual project description via Fable chat, I originally wasn't even going to model it in CAD and just build from the textual description. I had just seen someone else post success with FreeCAD + MCP and there were some measurements that were suspect... So I had it model it. It did pretty good entirely on its own but a wood bench is composed of effectively a bunch of differently sized rectangles oriented and put together correctly so it feels like this was an "easy task" for it to complete.

The other side of the coin was that inverted differential omni-wheel. This one I knew mechanically exactly what needed to happen and I wanted to model to make a parametric, re-usable model that could be dropped in to a bunch of different designs. This one needs some complex gearing built which it actually mostly did pretty well. Where it kept failing here was weirdly in the motor placement (it was a stock placeholder mesh of the motor part), it had a sled mount with screw slots to allow a little adjustment of the meshing of the final gears. Over half of my time working with the model was simply trying to get it to point the motor model in the correct direction, fully seated in the sled, with the screw slots accessible and physically not colliding with anything. The next hardest part was composite/boolean joined parts, specifically a couple of gears that are a combination of crown and pinion gear, it couldn't seem to comprehend that adding the two physical objects together wouldn't remove excess material from around one of the sets of teeth.

Kind of a ramble, but let me know if there are any more specifics I can provide.

I've also had really good success with FreeCAD, and the FreeCAD MCP server. So far I've had it design and build a custom workbench (woodworking project w/ cut list, assembly and hardware instructions, tool and part check lists) as well as a omni-directional differential style robot wheel module for a 3D printing.

I will say the differential wheel module was on the edge of what it was able to do. I had to bump it up to Fable and if I didn't already have a clear model in my head of what I wanted I don't think it could have pulled it off.

I don't think it's appropriately addressed. The apple ecosystem has really good syncing between devices but only when they share an iCloud account. What happens when your work mac isn't permitted to use your personal account? What happens when your company issued password manager and your personal one fights? What happens when your company issued device is an Apple device AND you have to use a mandatory password manager from the company?

The problem to me is that passkeys assume the user owns the device and CAN choose to make a single store authoritative for their credentials. The only professional environments I've encountered that haven't/wouldn't hit this issue are ones small enough that they don't really have device or security governance in place which you can't use as a secure reference baseline.

Huh. Well, federal court so national precedent against creating accounts to scrape registration only content. Probably a good thing IMO. I would put money on the public statement by a LinkedIn official saying "Your profile is yours" will bite them in the future as they certainly don't treat it that way.

Internal costs are rising directly due to AI spend, we're still figuring out cost control and budgeting around this but are expecting this to basically become a new per-employee cost to factor in.

I would say internal morale and productivity has improved. We have explicit policy guidance and a general collective distaste of pasting AI responses which is what usually sours me on interactions with people that are fully gargling the kool-aid. My CI experience has gone down from ~1.5 hours to less than five minutes and I can actually run our entire platform locally on my laptop again. My personal experience has improved at least.

Externally, you're probably right that we're not really capturing user sentiment well about how our company's changes are impacting them. I can concretely say that there were several customers that were disappointed that our public change log notes have been about the same length. They were expecting us to go through our feature backlog faster because we we're starting to use AI.

On the other hand, we have had fewer outages, reached our fifth 9 consistently over the past four months which is a new company record in the 11 years its been tracked. We have had fewer support tickets, our customer retention rate is the highest it has ever been in the company's history, and we're growing our internal teams. Based on the evidence I have available to me, I would say that we're healthier than before both from a user sentiment perspective and an internal morale perspective.

A good chunk of what my company has been doing with AI falls into either burning down our known tech-debt and "easy wins" that no one ever had the bandwidth to approach... And improving / automating our processes. The former is having a direct and meaningful impact on the quality and availability of our services.

Our QA, formerly a fairly frequent blocker of all our releases, are doing more in-depth reviews and catching issues earlier in our release process. They have become unblocked to the point they are actively chasing down work that starts to slip.

We have cleaned up and tuned both our security alerts and operations logs and improved our tenant isolation in our service in a way that makes customer and formal audits SIGNIFICANTLY easier.

We're setting ourselves up for faster human development of the hard-things. Our development environment and infrastructure are faster, cleaner, more auditable processes, and cheaper overall to operate.

These fixes mostly don't show up in our product change logs, and definitely don't fall into "new features". It would largely be invisible to the outside world, but our costs are going down (though to be fair, not offsetting the spend on AI to date), internal productivity has improved, operational incidents are down, and customer satisfaction is up.

I have a kid with my partner and I'm still not sure it was the right decision for me. They do become a full-time job in their own right and there is a whole set of skills that you can kind of learn through books but is going to be unique to your little one.

The first couple of months were especially hard as they don't have much of a personality, don't really react much to the world around them, they're just feeding, sleeping, and growing. Starting at the beginning of the third month for me it started changing and every baby is kind of unique.

I now _detest_ most of the books/articles/child-rearing advice I've encountered. Most of all of their content is fluff to pad pages, a blog roll, and are frequently superseded by more modern research backed evidence. There is a lot of toxic advice in those books and systems as well, usually for parents that are prioritizing restoring their prior social lives over the well-being of their children.

It's time-consuming and exhausting, but I haven't found it to be _hard_ yet. Waking up in the middle of the night and giving your kid a hug and maybe reading to them a bit when they have a nightmare isn't a challenging task. The challenging tasks are exceptions (surprise medical issues, your daycare closing down with two days notice, etc). The baby and child industry put a cost premium on the clothes targeting them and they don't last very long due to the child's growth. I _saw_ this coming but not to the extreme it has actually manifested and you'll find it replicated in every child/baby sized item.

There are a lot of surprising upsides and my motivations have definitely changed since we had a kid. I will flip a mountain to see my kid smile and laugh, it hits deep. I enjoy spending time with her and I'm terrified of missing her achievements and milestones. When she comes to me and asks me to show her how something works, wants a book read to her, or just wants a hive five I get a double whammy of that serotonin once for being a Dad that she wants to engage with and once for getting to spend time with her.

I still, and unfortunately kind of frequently, miss what I have effectively sacrificed to have her. I am slower at staying on top of emerging trends, hobbies have been left by the wayside that I did for decades, I gave up running conferences and with it I drifted away from whole treasured social groups. None of this was really a surprise, but I didn't realize how much what I gave up was really part of who I was.

You should go read the actual technical reports of the incidents and the follow on reports about the capabilities of smaller models in similar kinds of environments. This isn't new. The things exploited are still pretty basic in old and poorly maintained software or in gaps in architecture that were intentionally poked against security policies.

Are the findings valid? Yeah they're still doing security and they're still finding real zero-days. I think the internet is going to be bleak not because these models can ALL do basic security research but rather that the baseline quality of all deployed software is so low.

Ah yes let the FUD continue. This is a real problem but so far not nearly as severe as any of the marketing has made it out to be to the overall detriment of everyone including these companies announcing these scary capabilities. These announcements always included half hearted attempts at security layers which has now been demonstrated to benefit attackers more than defenders.

I wish I had a real solution to this beyond a dark age of the Internet where people have to finally come to terms with the general poor quality all modern software tends to normalize at.

What you're describing is a harness implementation detail not something specific to the MCP protocol or even how skills behave. Both of those are effectively providing the same level of information to the harness. Harnesses have traditionally (and still generally do) exposed enabled MCP servers and their actions ahead-of-time in the system prompt directly instead of doing progressive disclosure.

Really not trying to be cheeky... but why? Who is the audience here? I can see maybe academics with small grants and want to do the absolute minimum spend on compute... But that is an audience you will have to fight for every cent.

This doesn't solve or provide guidance for the subtle problems in these otherwise opensource solvers... The first example requires the client to manually disambiguate equivalent variables to get a stable solution... Sure that's a pretty common problem everyone working with optimizers should be familiar with but they're also one of the hardest things to track down in a complex derived model.