shrike's recent activity
-
Comment on Moltbot personal assistant goes viral – and so do your secrets in ~tech
-
Comment on The downfall of OnePlus will be studied | The "enthusiast brand" arc in ~tech
shrike Link ParentYea, I have enough M-series Macs for generic computing needs, I'm specifically looking for something that can do Linux gaming. The FW desktop has the added fun factor of being the only non-mac...Yea, I have enough M-series Macs for generic computing needs, I'm specifically looking for something that can do Linux gaming.
The FW desktop has the added fun factor of being the only non-mac machine that can actually do local AI as well as being decent for gaming.
-
Comment on The downfall of OnePlus will be studied | The "enthusiast brand" arc in ~tech
shrike Link ParentIf the Steam Machine doesn't pan out (too expensive or some other non-software issue), my next computer will be a Framework Desktop or their laptop with the GPUIf the Steam Machine doesn't pan out (too expensive or some other non-software issue), my next computer will be a Framework Desktop or their laptop with the GPU
-
Comment on Moltbot personal assistant goes viral – and so do your secrets in ~tech
shrike Link ParentSkills don't have to be massively complex. I'm just building a skill for work that just tells the whatever agent how to use logcli to access our internal Grafana and check the logs for whatever...Skills don't have to be massively complex.
I'm just building a skill for work that just tells the whatever agent how to use
logclito access our internal Grafana and check the logs for whatever project its working on.Yes, it can figure it out by itself, but it'll take a few tries since our specific setup needs some extra tweaks. With a skill it can just basically copy-paste commands from the skill files and get to work.
For my personal stuff I have a code analyser skill that uses Python + tree-sitter to parse a codebase and look for specific (AI-induced) shitty coding patterns - again something that agents can do without a skill, but the scripts provide specific and easy to parse data for the Agent and save a ton of tokens.
-
Comment on Moltbot personal assistant goes viral – and so do your secrets in ~tech
shrike Link Parent(Disclaimer, haven't used Openclaw, too creepy. But I DO know the tech behind it) This is exactly what Apple promised/teased us with AI Siri, but completely failed to deliver. With Openclaw you...(Disclaimer, haven't used Openclaw, too creepy. But I DO know the tech behind it)
This is exactly what Apple promised/teased us with AI Siri, but completely failed to deliver.
With Openclaw you can pretty much chat with a personal assistant who has the capablity to actually do other things than set timers and call people. The way it differs from "just chatting with ChatGPT" is that it can run on the background "autonomously" (basically it wakes up every X minutes to check a tasklist).
You can tell it something like "I'm at the office every Monday and Thursday and I use the D line on the train, notify me of any disruptions", then it'll make a task for itself to check the train schedules and will ping you on any communication method(s) you've given it. In its simplest form it'll just open the schedule page every few minutes near your departure time, read it and check the shedule.
It can also, if given the ability to, check your calendar and not notify you when your calendar says "vacation". It can also write a tool (a piece of javascript code) that will check the time tables more efficiently using an API. And depending again on a bunch of LLM mumbo jumbo, it might even look at the upcoming weather, past history of train disruptions and tell you on Sunday night that "It's going to be -29C tomorrow morning, which historically causes disruptions on the train schedule, would you like to wake up earlier?" Or it might read the news and see there's a transport strike that will affect your specific line.
Etc. etc.
It's not scifi stuff, we have the tech for all that today. Making it privacy-preserving is the only major hurdle. Openclaw just shoves all of your private info at an Anthropic API, which is ... sheeesh.
Apple's Private Compute is secure enough for me to trust it with calendars etc, if I want my bot to have access to my health data it has to be fully local, and the compute power just isn't there yet.
And yes, all of that can be "just an app", but with the agent in the loop it can handle the fuzziness of real world and incomplete data quite a bit better.
-
Comment on Why everyone is suddenly in a ‘very Chinese time’ in their lives in ~tech
shrike Link ParentSomeone said it well: Canada chose Lawful Evil instead of Chaotic Evil. Both are evil, but at least China is predictable, they don't do foreign policy via social media posts on a whim or due to...Someone said it well: Canada chose Lawful Evil instead of Chaotic Evil.
Both are evil, but at least China is predictable, they don't do foreign policy via social media posts on a whim or due to personal vendettas.
-
Comment on I recently finished the Cradle series by Will Wight and have post series depression. What shall I read next? in ~books
shrike LinkStarship's Mage isn't quite on the level of progression as Cradle is, but I've liked it for the 15 (out of 17) books I've read. It even manages to carry on when the (original) main character is...Starship's Mage isn't quite on the level of progression as Cradle is, but I've liked it for the 15 (out of 17) books I've read. It even manages to carry on when the (original) main character is demoted to a side character for story reasons - having built up a decent stable of secondary PoV characters.
It's magic - but in space. Good stuff :)
It goes from the MC being a basic jump mage to
but unlike "proper" progression fantasy like Cradle, there's always a price that will be eventually paid.Minor spoilers to a singular event around book 4-5
Stopping a full-on orbital bombardment -
Comment on I recently finished the Cradle series by Will Wight and have post series depression. What shall I read next? in ~books
shrike Link ParentI'm currently on book 2 of Beware of Chicken, 50% because of Baldree and 50% because I wanted something cozy to listen to. Haven't been disappointed so far, it's exactly what I wanted =)I'm currently on book 2 of Beware of Chicken, 50% because of Baldree and 50% because I wanted something cozy to listen to.
Haven't been disappointed so far, it's exactly what I wanted =)
-
Comment on Is it possible to live without WhatsApp? in ~tech
shrike Link ParentIronically of all the main non-video social media sites (twitter, insta, bsky, mastodon) Threads is the least bad for just randomly discovering stuff. I think it's mostly because it has pretty...I never used Threads
Ironically of all the main non-video social media sites (twitter, insta, bsky, mastodon) Threads is the least bad for just randomly discovering stuff.
I think it's mostly because it has pretty robust blocking tools and the algorithm rewards interaction, but in a different way than Instagram. The community also has a "block and move on" mentality, if someone is being an ass, you block them and go about your day, don't start arguing. If someone is dicking around in the comments of your post, you can block them so that nobody can see their messages under that post, cleaning up the interaction a lot.
As for the algorithm: On instagram your first 30 minutes or so decides whether your post will be on your follower's timelines, so "professionals" have upvote rings where they like, comment/share each other's posts just after they are posted, boosting their visibility initially, cheating the system.
Threads seems to require an account to have either actual interaction (reply to messages that aren't your own) or massive virality. Accounts that just post, but never reply get buried unless you specifically follow them.
It's still Meta, but not shit. Which is really weird for me, as FB and Instagram both are completely unusable.
-
Comment on Is it possible to live without WhatsApp? in ~tech
shrike Link ParentThe nerdier folk are, sadly, easier to bring to Discord or Telegram because both support bots natively. Signal and Whatsapp don't, which leaves a lot of fun out you can add with all kinds of silly...The nerdier folk are, sadly, easier to bring to Discord or Telegram because both support bots natively.
Signal and Whatsapp don't, which leaves a lot of fun out you can add with all kinds of silly self- or premade bots.
The only ones in my circle who are Signal-only are the hardcore privacy advocates, I've yet to find any group chats on Signal, it's all 1on1.
-
Comment on The truth about AI (specifically LLM powered AI) in ~tech
shrike Link ParentEU would definitely pass these regulations, even if it means shooting themselves in both legs. Russia won't, India won't, USA won't because basically all of the major AI subscription services are...EU would definitely pass these regulations, even if it means shooting themselves in both legs.
Russia won't, India won't, USA won't because basically all of the major AI subscription services are American.
China ... they're betting heavily on a "here's an open model, run it yourself" -apprroach. Mostly to annoy American companies who are subscription based and it's a lot harder to regulate what people do with their own models with their own hardware.
-
Comment on The truth about AI (specifically LLM powered AI) in ~tech
shrike Link ParentAt this point regulating AI is about as easy as deciding, as a society, to dismantle all nuclear weapons. It only works if EVERYONE does it - and that's not happening, sadly.I disagree that it's unavoidable. We could make a different choice as a society. We could regulate this shit.
At this point regulating AI is about as easy as deciding, as a society, to dismantle all nuclear weapons.
It only works if EVERYONE does it - and that's not happening, sadly.
-
Comment on Statement from Mozilla's new CEO in ~tech
shrike Link ParentOptional and the possibility to use a local model.Optional and the possibility to use a local model.
-
Comment on What's your favorite RSS reader? in ~comp
shrike Link ParentI can recommend NetNewsWire. It's free, has been free for decades and supports FreshRSS.I can recommend NetNewsWire. It's free, has been free for decades and supports FreshRSS.
-
Comment on New X/Twitter feature revealed many MAGA influencers to be foreigners in ~society
shrike Link ParentIt's in their playbook: https://en.wikipedia.org/wiki/New_generation_warfare They figured out that just generally destabilising everything and making normal people argue about non-issues on social...It's in their playbook: https://en.wikipedia.org/wiki/New_generation_warfare
They figured out that just generally destabilising everything and making normal people argue about non-issues on social media is an easier way to Make Russia Great Again than just outright war.
I'm 100% convinced that pretty much every major moral panic in "the west" has been either initiated or massively amplified by the Russian Online Troll Factory.
They're super active on every single social media video about the Ukraine war on anything to do with EU politics in general, pretending to be just normal Europeans super worried about "unelected" officials in the EU "warmongering" and wanting a war with Russia - a perfectly peaceful country. Within 1-2 minutes of any such video being posted, there's a clear template argument with 50+ likes under it already.
-
Comment on Bun is joining Anthropic in ~tech
shrike Link ParentCrush is definitely prettier, but Claude is more usable by far.Crush is definitely prettier, but Claude is more usable by far.
-
Comment on Bun is joining Anthropic in ~tech
shrike Link ParentGood thing Google is only the world's biggest ad company, so they're cool. Right? =)Good thing Google is only the world's biggest ad company, so they're cool. Right? =)
-
Comment on Building the perfect Linux PC with Linus Torvalds in ~tech
shrike Link ParentHe has the exact same attitude towards AI as me, something about the practical Finnish mind or something =)He has the exact same attitude towards AI as me, something about the practical Finnish mind or something =)
-
Comment on Are there any current Kagi extended trial codes? in ~tech
shrike (edited )LinkThey just sent a bunch for the holidays, I've got 3 I can share (also sharing elsewhere, so the reply count here might not be accurate to remaining codes :) ) Easily the best search engine and...They just sent a bunch for the holidays, I've got 3 I can share (also sharing elsewhere, so the reply count here might not be accurate to remaining codes :) )
Easily the best search engine and service I pay for. It's in the top5 of stuff I'll unsubscribe from last.
EDIT: All codes sent
-
Comment on Swedish death metal band Arch Enemy has parted ways with its lead vocalist Alissa White-Gluz – announces upcoming release of her solo album in ~music
shrike LinkJust saw them live for the first time a few weeks ago. Couldn't have picked a better time it seems =)Just saw them live for the first time a few weeks ago. Couldn't have picked a better time it seems =)
The basic idea with skills is that... (Analogy time!) instead of the agent having a workshop with 1000 tools (MCPs, massive AGENTS.md), giving it analysis paralysis where context fills up and the agent legitimately gets confused which tool to use, it has a set of books on the wall - skills.
When it comes in to the workshop (fresh context) it reads the spine of each book (skill name) and the blurb from the back (skill description, a sentence or two), trying its damnest to remember them all.
Now if it needs to create a PDF it remembers that it saw a book about it when it came in, grabs it off the shelf and reads it. Now it knows kung fu - or to make PDFs.
Or the user asks it to deploy something to Cloudflare, again. Skill!
Although it seems that the current generation of agents prefers stuff to be in AGENTS.md with references to documentation in there: https://vercel.com/blog/agents-md-outperforms-skills-in-our-agent-evals
All of the major players are pushing for Skills, more so than MCPs in some cases, so that issue will most likely be managed or mitigated within weeks though.