Alex Levberg

Quickly activated @SimpleAnalytic to see if I get any traffic on todays launch. Mote than happy that over 100 of you checked it out 🫶

@eddiejaoude Okay fair point maybe I just got a bad substack experience. From 25.000 Views I got 125 Subs. But could be my content and not Substacks fault

@alexlevberg I still feel those messaging apps are for friends and family, I don't want big groups or broadcast channels on there - maybe that is just me

You don’t control what content your community sees, the algo does! YouTube -> algo Twitter -> algo Instagram -> algo TikTok -> algo LinkedIn -> algo … Newsletter -> YOU Time for me to spend more time on Substack, what about you?

@eddiejaoude I thought of more like a Channel (Broadcast) as a Newsletter alternative, not a Group

@alexlevberg I never join any groups like that, my push notifications are precious

Hi @helloitsolly , would @SenjaHQ be open to offer the Growth Marketer Role as a Freelance position?

@init_bobjames You can run it on iphone and Android, just search for "Google AI Edge Gallery"

@alexlevberg oh my days thats so cool 🙌 Google Tensor G3 1x 3.0 GHz Cortex-X3, 4x 2.4 GHz Cortex-A715, 4x 2.15 GHz Cortex-A510 GPU: ARM Mali-G715, 890 MHz Cores: 10 RAM: 8 GB, 4266 MHz might be better spec than my laptop 🤣

Gemma 4

@OliDietzel @itsolelehmann normal RAM is slower than Apples Unified Memory, you would need VRAM from a GPU

@alexlevberg @itsolelehmann Would go with a ryzen h255 mini pc 96gb. 780m igpus have full um under Linux and Vulkan.

it's interesting how the math to buy a computer changed it used to be: - it is fast enough to run my apps now it's - how much money in token can it save me running local models for me so buying a computer for 5-8k can be cheaper long term than paying 2k-3k + token

@DruRly But whats the USP, most folks could built it themselves with @openclaw

@itsolelehmann now your new goal when building will be to secure your son's future

so thankful my son came to this earth healthy everything else just isn’t as important

@thekitze maybe try: TypingMind from @tdinh_me

is there a better lm studio alternative for chatting with local models? who the goat?

@kyjus25 @brobro @thekitze thx 😊

@brobro @alexlevberg @thekitze It is LM Studio

bro i was fkn sleeping on local models. these mfs have vision now?? meaning i can run periodical security camera checks .. or take a screenshot of my work every minute and keep a memory log of what i'm working on... FOR FREE???? so many fucking use cases for this 🤯

@eddiejaoude @init_bobjames @jcubic @Stew_SoFresh looks like a great community, the last place, safe from AI Slop

Great to geek out with @init_bobjames @jcubic and @Stew_SoFresh 🔥 Time went so fast, we talked about content creation, interviews and of AI, plus a bit of text editors LOL Thanks Bob for organising! Hopefully see more of the tech community at the next community call ☎️

🦞Ollama's cloud is one of the best places to run OpenClaw. $20 plan is enough for most day to day OpenClaw usage with open models! To make the switch, all you need is to open the terminal and type: ollama launch openclaw Choose a model: kimi-k2.5:cloud glm-5:cloud minimax-m2.7:cloud If you are affected, Ollama welcomes you!! ❤️

@thekitze None

what are your usecases for openclaw/hermes that you cannot do with codex / claude code (local or remote via phone) I am legit asking and not trolling.

@gregisenberg So most people fail with distribution after they built something or fail with the idea in the first place

the gap between "i have an idea" and "i shipped a product" just got so small it's basically not a gap anymore for anyone, anywhere

@nanacmg47 @eddiejaoude you can use kimi for free with Nvidia or 20per month over ollama

@alexlevberg @eddiejaoude Hi how much u pay for kimi per month, im searching for claude alternative, i have claude max 5x but keep reaching limit 🥲

Claude Sonnet 4.6 > GPT 5.4

@itsolelehmann Think i will have to pull the trigger and go with Perplexity Computer. Never had a real use case for openclaw but perplexity seems to be the thing i can really use

so you're telling me perplexity computer can now: > draft your federal tax return on actual IRS forms > catch errors in professionally prepared returns > and update itself as tax laws change? they even tested it against real tax attorney prepared returns and found a 67% error in overtime deductions which saved the filer thousands of dollars he would've otherwise never claimed accountants are officially cooked lol but my hot take is that this is a GOOD thing the boring part of accounting (read docs, apply rules, fill forms) was never the valuable part anyway. the valuable part = judgment, strategy, knowing when to tell a client "do this" or "don't do that" ai just freed up accountants to do the work that actually matters. the ones who adapt will be worth more, not less