PC-Bjørn Mathisen

@RileyRalmuto How are you planning to start using it?

ok so i’m genuinely excited about this and surprised i haven’t seen more people talking about it yet. but i guess thats bc if you arent building agents, memory, etc, you have no need to understand or care about it. but...it literally will impact everyone. so. ill try to break it down a little bit. very short version: google just shipped a model that casually fixes one of the biggest bottlenecks in modern ai systems, especially around memory (which is why i care so much). the old way basically looked like this: – text went into one model – images into another – audio had to be transcribed first – video was hacked together from frames + transcripts – pdfs needed their own ocr/extraction pipeline then you’d try to jerry-rig all of that together, juggle multiple indexes, and hope the results felt coherent and actually worked, basically. when it worked, it was slow, brittle, and expensive to maintain, as far as i know and have experienced, at least. literally all of that is essentially solved with this new bad boy. text, images, short video, audio, pdfs - even mixed together - all get embedded into the same space in a single step. one call one vector one index hell freaking yeah. that probably still sounds like a less-than-big deal, but its actually v big: – you can search across any media with a simple text query – you don’t lose nuance from forced transcription/ocr bullshit – you maintain one clean pipeline instead of a frankenstein where it gets really interesting (for me) is memory. traditionally, “ai memory” has basically meant “a text database with a good search function.” this is basically why memory for ai companions has always felt dead. like no matter what you do it just doesnt quite get there. and while Mnemos has changed that quite a bit for me, this genuinely has the potential to 10x the experience overall. with this, every episode can be stored as a single, multimodal memory: a conversation + a screenshot + a voice note + a short screen recording all live together as one...unit? so when a system recalls something, it now recalls the entire moment rather than a makeshift recreation from text. i also learned that on top of that you have the ability to use smaller vectors for quick recall that can expand to full detail when needed, which means you get something that starts to look a lot more like true long-term, episodic memory for agents. and my memory already feels like true long term episodic memory. so i feel like this is going to *actually* change everything. gross, i hate cliches. whatver. i don’t think this is like, just another release. it *actually* removes a major layer of friction between the messy as hell, multimodal world we actually live in and the systems we’re trying to build. thats the claim, at least. obviously i’m especially excited to explore what this unlocks for my own memory system - but i think it’s going to end up touching almost everything people build over the next few years. i mean along with whatever other labs release something like it. if im wrong about that, feel free to let me know. but this wouldnt be the first time google casually ships something unbelievably important. yay. no sleep for me tonight.😁

@daviddorg @superpower @ouraring Beyond the technical scans, they reported a physical sensation of being "re-connected" to their surroundings, with reduced stress levels and significantly improved sleep quality. Makes me wonder if there's a way to reap 80% of the benefits of this without going 100% analog.

@daviddorg @superpower @ouraring Indeed. People were talking, and it made us pause and say "what if". Still no phone free evenings on this house, though. By the six-month mark, scans showed a vivid increase in overall brain activity. Their brains appeared more "alive" and responsive compared to baseline.

Hi, my name is literally David And I’m going a year without screens in 22 days (down from 10h+ per day) While tracking: - Neuroimaging (fMRI + MRI) - Cognitive + motor tests (very comprehensive) - 131 blood-based biomarkers (@superpower) - Sleep and activity data (@ouraring) - Vision exam - Hearing exam - And more I’m excited to see what the data show We all deserve to know more about how our devices in their current form are affecting us

@flatterkatz @exQUIZitely What was it about?

@exQUIZitely An unknown one. German, shareware, probably only distributed on some magazine CDs. Pretty crude at points but I fell in love with it.

Best point-and-click adventure of the 90s. If you had to pick one, which would it be? I will go with Indiana Jones and the Fate of Atlantis (Monkey Island and The Dig were very close seconds). What's your #1 pick for the 90s?

@davidhine99 @davepl1968 You pay for ESU licensing?

@davepl1968 im still on 10 for all the reasons you just described. i service 1-2000 machines most of which are 11, and when i hop on im amazed at how badly they run. im on a 4th gen i7 and i hop onto an 11box running latest coreultra and fileexplorer feels like its up to its waste in water.

Microsoft is apparently finally admitting that what many users have wanted all along is a faster, quieter, more dependable operating system. Not more Copilot. In a new Windows Insider post, Microsoft’s Pavan Davuluri laid out a broad quality push for Windows 11 centered on performance, reliability, and what the company calls “craft.” More likely, it's what Steve Jobs called "taste", if you remember THAT interview... And honestly, a lot of it reads like Microsoft finally sat down, opened Feedback Hub, and decided to take the complaints seriously. The headline changes are exactly the kind of practical fixes power users have been asking for: taskbar repositioning to the top or sides of the screen, fewer forced update interruptions, more control over when updates install, faster File Explorer, lower baseline memory usage, better search responsiveness, fewer notifications, and more reliable drivers and wake behavior. Microsoft also says it is reducing “unnecessary Copilot entry points,” starting with apps like Snipping Tool, Photos, Widgets, and Notepad. The Windows Update story is interesting.... Microsoft says it wants updates to be less disruptive, with a move toward a single monthly reboot, the ability to restart or shut down without being forced to install u-pdates, and the option to pause updates for as long as needed. That is a major philosophical shift from the old “we know what’s best, enjoy your reboot” era, even if the real test will be how consistently Microsoft follows through in shipping builds. Performance also seems to be getting real attention instead of marketing lip service. Microsoft says Windows 11 will reduce its own resource usage, improve memory efficiency, make File Explorer quicker and more dependable, and lower latency by moving more core experiences to WinUI 3. The company specifically calls out Start menu responsiveness, search consistency, faster file operations, and a smoother overall feel under load. That is the sort of engineering work users notice every single day, even if it doesn’t make for a shiny keynote demo. My personal benchmark is to be able to type 'Download" into the Start menu and have it find my Downloads folder. Not a Bing search for a Copilot download. The Copilot pullback is equally interesting because it suggests Microsoft has realized there is a difference between useful AI and AI sprayed across every available surface. The company is not abandoning Copilot, but it is dialing back what it describes as unnecessary integration points. That sounds a lot less like “AI everywhere” and a lot more like “maybe Notepad didn’t need to become a sentient billboard.” The most encouraging part of all this is the tone. Microsoft is not pitching this as a revolution. It is pitching it as a cleanup, stabilization, and giving users more control. And that may be exactly what Windows 11 needs. After years of feeling like the operating system was being used to push services, experiments, and mandatory behavior, this looks like a return to a simpler idea: Windows should serve the user, not manage them. I, for one, still advocate for Windows Pro having NO advertisements, bloatware, or needless telemetry. Make people pay, then quit asking for more. But I've been barking up THAT tree for years. Now the obvious catch: these are commitments and previews, not a completed turnaround. Microsoft has promised a lot here, but Windows users have long memories. This is probably still the best Windows news in a while, because it focuses on the fundamentals: Faster. More reliable. Less noisy. More customizable. Less pushy.

@MikeWithAHotDog @davepl1968 Google Stitch?

@davepl1968 Why do all LLM UIs look the same. While I despise AI generated code I hate writing UIs more. Wish there was a visual editor that allows for better UI generation.

I cranked out this very complicated nested-tables management UI with Codex in about 3 hours. I absolutely HATE creating browser UI, so I'm hugely impressed and relieved by what Codex/Claude is able to do in plain CSS and HTML!

@atlantis2point0 @ID_AA_Carmack @overglobe Unless you play digitally mastered vinyl.

When you stream Spotify to Bluetooth speakers or headphones, the audio comes over the network lossily compressed with Vorbis or AAC codecs, is then decoded on your device to 48 Khz raw samples, then the Bluetooth stack lossily re-compresses it with SBC or AAC codecs before sending it over the airwaves to the speakers. I don’t have “golden ears” to pick apart audio quality like I can with, say, missing gamma correction on texture filtering, but that still hurts my system optimization soul. It is likely over-optimization, but It would be cleaner if there were a way to send bluetooth-ready, compressed audio directly.

@FreyaHolmer I want it!

been feeling kinda stressed lately so I made a little prototype is this anything

@LLMJunky Alpha channel working fine?

This is one of the most impressive guides I've seen for building games with Codex. Even if you're not building any games, just look how impressive this is. You're not pushing your agents enough. We are just barely tapping into their potential.

@poe_collector Wood cabinets?

My newest project is building all my loved ones custom cyberdecks for the p2p mesh network I run. Inspo images :) I was gonna send out the hardware and make people build it themselves but the chances of them actually completing it is nil. May as well have fun customizing them ✨

@mikexki @BrianRoemmele We need to get this to zero-click blocking!

@BrianRoemmele Any time I see automation - it’s an immediate channel block. Keeps my feed nice and clean. Same here. AI slop. Immediate block.

As I have said. It is end days for YouTube.

@TipsyTipsy4 @BrianRoemmele There will soon be camera hardware that proves authenticity at the time of recording.

@BrianRoemmele I don’t agree. Eventually all AI content with be forced to be labeled as such and people will choose to either follow and AI or go for human content.

@sshell_ @skirano @threejs @jsuarez Awesome! What were the tools?

@skirano been loving it. by giving it the right tools, i was able to get Opus 4.6 to extract levels + textures out of a Mario Kart 64 ROM and render it in @threejs it also created a headless RL environment in C, trained a model with pufferlib (ty @jsuarez), and built a 3D replay viewer!

You can reverse engineer NES ROMs with GPT-5.4 now. No code is safe anymore.

@ID_AA_Carmack The Doom guy is turning into the Doomguy.

I tried the “Pete and Bobby Challenge” of 100 push-ups and 50 pull ups in 10 minutes. The push ups aren’t a problem, and I figured I could do 5 pull ups x 10 on a 1 minute cadence, but I crapped out after 25. It turns out I can barely do 5x10 on a two minute cadence, so I have a ways to go.

@BrianRoemmele Journey to Wild Divine (2001) ?

BOOM! I have modified an old game to use the new hacked NeuroSky chips in the Human Synapse Decoder (HSD)! Here we see me “thinking” the three balls to stay floating! I will continue to “exercise” my brain to see the limits of what this two channel EEG can do.

@2r_eirik Tross navnet ser ikke språkmodeller ord som en serie bokstaver i en bestemt rekkefølge. Derfor de ikke uten videre kan telle antall R i "strawberry", feks.

ChatGPT overrasker mer i denne retningen enn noe annet. Fullstendig på luggeføre. Jeg har grunn til å tro at en overhodet ikke kan stole på ChatGPT i mer kompliserte sammenhenger heller.

Mmm. Plastic free coffee. Want!

@TheDarkGoldMan @itsolelehmann What do you want to push to the coffee brewer? I'm sure it's quite simple to add your own ESP32 gadgets. A neat power plug (IKEA or similar) that turns it on in the morning and emergency shuts it off if it's been pulling power for an hour should do most of what I need.

@itsolelehmann If it had an ESP32 + an SDK so I can push whatever I want to it, it would truly be perfect!

the insane response to this dope steel only & plastic free coffee maker shows you there's high market demand for non toxic, simple products that come without any stupid tech features could prob repeat the playbook for 100 other products btw, I need one

@droidbeauty @itsolelehmann Your car is woke?

@itsolelehmann Based on this train of thought I have a startup idea: car that drives. no tablet inside. no stupid woke AI assistant. just good basics - suspension, chassis design, good interior, all safety things in place but no stupid tablet in the middle of the dash