I’m an ML and data engineering consultant, so whatever time I spend working is something my clients have determined they need enough to pay my hourly. Whether the market needs their products is tbd. But there’s not a shortage of demand for work in this space right now.
Sometimes my job feels an awful lot like detangling thread. Keep pulling until you reach some friction, isolate the individual components, figure out which pieces are causing the issues, correct. Repeat. Deploying new services (vs features) is often like this. I can’t say how many times a day my colleagues and I say, “Yes!! A new error”. Experience just makes it easier to identify the issues and isolate the pieces faster. And to not get discouraged or stressed by errors. Expect them - you’ll be a much better tester and will find bugs earlier (and hopefully before prod).
?cid=2154d3d78iyxc85jpsbflnkacmdreoigzkf99gjvhcmfqpvp&rid=giphy.gif&ct=g
Yeah! Naive Bayes has a great history in email spam detection (and honestly I love it for a lot of short text string classification in general). I think the approach will be an ensemble approach, one leg of which will definitely be utilizing bayes.
This is really interesting. I used to work for a company that operated as a data steward for hospitals to analyze their claims data, and there was always a lot of discussion about FHIR. I could definitely see companies like this being essentially private relay operators, with EHRs integrating with them. If there were a sufficient number of private relays with hospital contracts, they could create a pretty well distributed network that would enable me as a user with a log-in to an EHR, to instantly share my record with a different health system. A lot of the data “risks” of third parties holding medical data are already being taken today for the purposes of analytics and research. The same compliance rules and standards could apply to relay operators that accept this data. I’m envisioning a much harder barrier of entry for relay operators, being gatekept by those that health systems are wiling to write to (obviously an individual can do whatever they want with their own data outside of HIPPA standards). Super interesting idea.
Recipe? Looks like a similar fermentation process to yogurt? Can I use yogurt as a starter for it or is it a different culture?
Such a mood.
I’ve started scrolling with my left thumb instead 😂
Yeah, I love redis, too. Definitely my go to for centralized in memory caching, and I usually start with it for basic queues as well (since most projects I have already use it). There’s just a complexity point where rabbit starts to make a bit more sense. But they actually work really well together.
Yeah, a lot of people I know are directly impacted by this for exactly that reason 😔
You could accomplish parts of this with redis pubsub and other parts with streams, but rabbit combined the best of both of those thing with even more functionality.
One of the things I really enjoy about rabbit is the fan out pattern that it supports. One job can publish events that N number of other jobs can subscribe to. These consumers can be entirely decoupled, with multiple workers - each consumer can get the same events (redis streams support something similar, but they don’t support ack/nack and handling dropped messages isn’t as graceful as it is with rabbit) and a consumer can subscribe to more than one queue at the same time. It also supports routing keys that can make distributing messages even more configurable, as well as dead letter queues that can be used to route errored or timed out events to other queues that can have their own set of consumers. Instead of having to spin up resources in terraform for new queues/consumers, each job can self register with the rabbit cluster programmatically and configure itself with its own specific filters. It can be consumed from both in blocking and non blocking ways, and supports auto nacking with heartbeats to ensure messages don’t get missed.
Spent my day working on some data engineering pipelines that make use of rabbitMQ. It really is such a versatile queuing system when you need more than the simple queues/streams redis can offer and aren’t pushing super high volumes that require something like Kafka. One of the most configurable message brokers out there 👌
Yes - at its most basic level. But I also think there are tools that we can build that makes this easier, more granular, and improve the user experience. It could be labels that aid in discovery or labels that aid in avoiding hate speech, or even groups of labels that I might want to use if I’m working vs social vs doing research. My wishes and desires may not always be the same at any given point in time, or nostr gets so big, blocking every single troll becomes a negative and annoying experience. I just think we can improve the blunt instrument, while still maintaining transparency and choice.
I agree on the distinction between bots and scum, and scum’s rights to free speech, but I also wonder if there is a middle ground for users to make decisions about what they want to experience. I think this is walking a line, and nothing should be automatically removed. But, what if there were content labels that gave power to each person to opt in or out of certain “topics” or sentiments or xyz. This puts the onus on each of us to step outside of our own echo chamber and walk that line ourselves, and is really about developing tools that enable us to do that. Or do that in doses, as we see fit. I think transparency and high levels of configurability, if done well, could offer something valuable. I, as a user, reserve the right not to listen just as they reserve the right to speak
🤷🏻♀️
I think at the root of this, there is a lot of fear. Of the unknown, sure… but also of what humans have done with spirituality/surrender, and the ways something so fundamental has been used to manipulate and also to shame (largely through this belief that “knowing” is possible and that it’s rooted in the past). I think what you’re describing creates a space in between… a way to not “throw the baby out with the bath water”. I think if you can bridge that gap, there is a lot of value there. You should definitely write the book!
That’s true 😳 sometimes at night it’s so sad. Those bunny squeals 😭