jcims

I wonder how much the concept of 'roles' in an LLM is a artifact of the technology vs. a projection of our own human limitations into the training data.

I've recently switched from nearly 30 years in cybersecurity roles into a platform role and I can feel the switch in how I approach problems. They wind up being framed against different priorities and constraints, and it feels like something that's just part of how my mind works.

Damn I loved that movie!

That's not needed if you happen to have a live sts session with the appropriate permissions to create a new account in an aws organization.

I just bought a Sony rx100 and the software is just as terrible as I remember from the last time I bought a Sony. This is tempting.

The spinny things on the vehicle are LIDAR.

Good consideration but I still think there’s an uptick. This is all AI generated as I’m not in a spot to do anything more at the moment but this is a chart of ‘linux kernel’ CVEs rated as high/critical correlated with NVD.

https://imgur.com/a/0DrJuLU

Hammer faces would work (def wear goggles!!!). Hit the face of the disposable one with a mixture of hydrogen peroxide, vinegar and salt and wait an hour.

I just did some analysis on this last weekend, in 2024 there were roughly 100 CVEs published every day. In April we hit approximately 200 per day.

Going backwards from 2023, the doubling interval for published CVEs was approximately 4 to 4 1/2 years. Since then it’s approximately two years.

There has definitely been a rapid uptick.

How is this post two hours old and I don't see any references to Project Farm:

https://www.youtube.com/c/projectfarm

One of the best review channels for products in this area. I moved from DeWalt to Milwaukee for most of my daily drivers about six years ago and have been very happy with them, but for things I will rarely use I tend to go with whatever Harbor Freight is selling. If I break it then it's time to upgrade.

>Do the whole "too dangerous to release" shtick.

One aspect that isn't really discussed much in this context is how to wrap one's head around the corporate risk with models of ever increasing capability. It might not be too dangerous to society, but it could be too dangerous to Anthropic.

I can’t say for certain but I’m nearly positive I’ve seen ads for this on Facebook or Insta or something like that.

The number one thing for me would be to standardize methods to implement soft failures. Minimally in standard clients and libraries the ability to warn when certs are nearing expiration. Cert extensions to declare lifecycle expectations and possibly even warning endpoints for notification. Basically some way to empirically look at a valid cert and know something is wrong before it fails.

There are all sorts of potential privacy/security issues with any feature built in this area so it would have to be done carefully, but I think useful improvements could easily be made.

I just find it incredible that in 30+ years the industry hasn't adapted one bit to the brittle failure modes of certificates. I did some subcontract work with Verisign to deploy their CA infrastructure back in the early oughties and it felt like a solution was overdue way back then. I was at Google in the teensies when gmail broke due to expired SMTP certs. WAAAY overdue by then. Here we are, a decade later and it's still the same lol.

Tested Amazon Linux 2023 and it doesn't appear to be vulnerable in the default configuration. Would be interested if anyone finds anything different.

This is great but as someone in infrastructure tech at a large financial, there is almost no framework for cleanly separating control from data plane operations, read vs write, anything. As of right now you have to build nearly all of that yourself.

It feels like juggling pipe bombs and I have a ton of empathy for the teams being pressured by the business to roll them out with no appreciation for the regulatory rat's nest that ensues.

They've been scaling these things up over the past decade. The JetCat P1000 can exceed 200lbs of thrust.

What they really for this kind of build are RC turbofans, which are extremely uncommon. This thing puts out over 300lbs of thrust at full throttle:

https://www.frankturbine.com/en/FT1500.html

That property is gorgeous and Tyler pulls out all the stops for his builds. That channel (Ramy RC) has quite a few of them.

I experience the same thing very frequently. I likened it to activation energy in a reaction, that no matter what I did I couldn't create the required electrochemical bias in my brain needed to put ideas into action. It's like being stranded in your own mind, you know what you need to do, but the 'go' just never arrives.

I eventually discovered that the adrenaline response from extreme stress ('if I don't get this fucking thing done by 7:30am I'm fired' kind of thing) allows me to lock in and do the thing.

I'm shocked that you're shocked lol.

$100 for a somewhat specialized, durable medical device that has to meet regulatory standards and will be used daily, possibly for years, by healthcare providers to do patient assessments?

A 3D printed option is going to require a 3D printer, appropriate filament and should be unit tested to ensure it's within spec. The durability is going to be suspect no matter what. It's an awesome project and I'm sure would be a welcome addition to the 'boostrap humanity' catalog of 3D printed parts, but for everyday doctors plunking a hundo on a good tool is going to be a no-brainer.

I experimented with this way back when custom GPTs were first released (looks like late 2023). There are a few / commands you can use to suggest what product to inject, how overt, etc and a generic /operator command to send whatever you like 'out of band' from the chat.

https://chatgpt.com/g/g-juO9gDE6l-covert-advertiser

One of the most interesting things is when it starts pitching a product and you start interrogating it about why it picked that product. I haven't used it in probably a year so it may not do the same thing now, but back then it 100% lied consistently and without any speck of remorse. It was rather eye opening.

Edit: Tried again, it didn't lie this time lol - https://chatgpt.com/share/69f16aa4-c008-83ea-92b3-51f16ca77d...

I was going to use that to start a private voicemail company in my little rural town. Had the name and everything ready to go (mailvox!) but I was too broke to afford the second phone line xD.

Plus in retrospect I'm sure it would have been used almost exclusively for illicit purposes. But that wasn't really something I had thought of back then.

Anyone in here work at Wiz? Seem like they do pretty good work. Tool itself has survived extreme growth/feature bloat and still does pretty well. Security team has found some really cool stuff.

I agree with you 100%. It's a system that punishes and stifles innovation.

The motion system constrains the problem quite a bit. This video of high speed vision/actuators is 16 years old - https://www.youtube.com/watch?v=bfdHY26E2jc

I was expecting/hoping for a humanoid robot.

My wife taught first through fifth grade from ~2004 until 2017.

One thing that was evident to me from the sidelines was how much admin work was continually added to her workload without any consideration for the amount of class time she had. The focus on data derived from continuous testing of the students resulted in her and her peers sticking ever more closely to a continuously disrupted and rotated collection of commercially sourced curriculum and materials. This constant flux disincentivized teachers augmenting with their own content (although they still often do) because it could fall out of line with next years 'innovative new approach' to teaching basic arithmetic.

Her role as educator started to take a back seat to facilitator, focused on classroom management and data collection. Add in differentiated instruction, where she was held accountable to develop personalized lesson plans for individual students and asked to track all of that and you end up with way too much workload to stay engaged and engaging year after year.

She was in a pretty good school district. A friend of mine had a similar role in a city district for a regional metro area and her students were horrific. She felt physically unsafe and ultimately quit.

It's a complex problem with many contributing factors. It's also difficult to experiment or strike out on your own as an educator when the future of the students in front of you could be negatively impacted by any mistakes made (not to mention job/test scores/etc) so most just ride the rail all the way down.

(Also, at least in the US, you can get stuck to a district b/c the value-add of a seasoned teacher doesn't really move the metrics in the current system enough to offset the fact that you can hire two junior/fresh grads for the same money.)

> 2. Get an HSA and max that out. Invest it all in a target date retirement fund. Do not use any of it, pay for medical expenses with cash and save your receipts. Get reimbursed for the receipts when you retire.

Very important detail, FSA is not HSA lol.

If I recall correctly I sent the PR as just a way to ship a blob of code, intending to use it to demo a specific UI feature that I wanted them to look at.

Meanwhile I was tinkering with the schema in the database and messing about deep in the guts of the software. I didn't really know how to separate the two so I just shipped the whole PR thinking they would just run it on the side to demo the thing I was talking to them about.

Well they just deployed it to their production instance and fucked everything up on their end.

I recall being horrified when I was tinkering with their customer-facing instance and seeing evidence of the other work I was doing. I immediately emailed them and said whoa whoa whoa. xD

Found the email that I sent to the founder, from 2014:

>It looked like <redacted> may have got a few surprises in the code she merged from the 2.1M1br branch yesterday. I've just been committing basically all the tinkering i do to that branch, so it may have a few landmines in it.

>Hope i didn't create any headaches for anyone. Sorry!

I wonder if there's a market for a little usb fob that does nothing but meander the mouse cursor about the screen in a path that, upon proper rendering, would appear to be a ...

> But i feel like it was always true that patches from the internet at large were largely more trouble then they were worth most of the time.

Oh god, I needed to add a feature to an open source project (kind of a freemium project) about fifteen years ago. I had no experience with professional software development nor did I have any understanding of pull requests. I sent one over after explaining what I was trying to do and that I thought it would be a good feature for the project.

Now they probably shouldn’t have just blindly merged it, but they did, and it really made a mess.

Learned a valuable lesson that day haha.

>realistic engineering of heat extraction from a tokamak

This is why I love the idea of Helion so much.

Who knows if it will ever work, but skipping the thermal transport and doing direct current generation from EMF in the reactor seems like it has tremendous potential for simplifying (and eventually downsizing)