localsonly
506 posts
localsonly
@LocalsOnlyAI
Agency in the Agentic Era. If it doesn't run on your machine, you don't own it. Locals Only
Katılım Nisan 2026
298 Takip Edilen132 Takipçiler
In 12 hours we've had 323 people sign our petition to protect local AI.
Open Source must win, not because anyone else must lose. But so that we can all win.
Please help me get 10,000 signatures so that we can walk into the room and say people care.
righttointelligence.org
English
I make the philosophical case for tax fairness and taxing billionaire wealth.
I welcome an argument on the merits for this new social contract.
Read here:
open.substack.com/pub/rokhannaus…
English
Put this in Trump accounts.
@altcap
Andrew Curran@AndrewCurran_
OpenAI is proposing handing over a 5% stake to the Trump administration according to the Financial Times.
English
@morganlinton It’s a good top 3. Marin, Tahoe, Sea Ranch for me.
English
@LocalsOnlyAI Marin is my third favorite spot on the West Coast, first is Tahoe, then Yosemite.
English
I always tell people, if you move to SF, you will fall in love with Marin.
SF is a great city, but it’s hard to beat the beauty of Marin.
marisa@meshtimes_
sf, i’ve spent 7 years with you and you’ve always been so good to me. i’ll always always love you before leaving, there was one last thing i wanted to do and i did it 🥹
English
And then they realize they can’t use Chinese models for a ton of work flows and end up going local.
Rohan Paul@rohanpaul_ai
UBS says 60% of companies now watching AI budgets are moving to cheaper models and open-source Chinese models The pressure is coming from extreme bills, including users spending up to $35K/month, teams exceeding quotas by 200%, and companies cutting internal AI tools from 5 to 2. Companies are not abandoning AI, they are using model routing, which sends easy tasks to cheaper models and saves premium models for hard reasoning, code, and long-context work. Chinese open-source models such as Qwen, DeepSeek, MiniMax, GLM, and Kimi now fit the enterprise cost curve because they can be run locally or used through cloud catalogs. --- news .futunn.com/en/post/75068082/ubs-group-finds-60-have-already-started-curbing-ai-spending?level=2&data_ticket=1780870170397383
English
@blackwellboy @NVIDIAAI I have the Asus and would rather have them match, but the 4TB storage would be nice.
English
@LocalsOnlyAI @NVIDIAAI what do you have now? i like to match them but if i get a founders edition, its going to be stacked right at the top as the lucky 6th !
English
nobody tells you the first DGX Spark is a gateway drug. I’m 5 deep in a 3D printed tower. I’ve started calling them “the boys” like I’m raising them. I am not seeking help or wife’s approval. I am seeking a 6th.
@NVIDIAAI you beautiful bastards. Just need a founders edition.
English
@loktar00 Wonder if Asus price will drop back down to $3,500. Doesn’t make sense at 1TB.
English
@CardilloSamuel Been testing on 1 DGX for my company. Plan will be to role out 1 or 2 per team and minimize our API costs.
Talking to a smaller hedge fund later this week about a similar approach in their company.
English
just had a call with yet another company switching to local ai. they were using Qwen 3.6 35b served through my spark so they could test their pipelines through it and see if its a fit and what i heard in that call is pretty crazy and i feel it is something that most anti-local ai misunderstand.
what the customer told me is literally that they kept comparing the results between Claude, ChatGPT and the locally served Qwen 3.6 and largely preferred the results they were getting via Qwen - insane!
the plan is now to deploy a DGX Spark in their office for the MVP + a 6000 pro blackwell-enabled machine for training & later on they're looking to basically buy a DGX Spark per customer they have so they can serve each customers an optimized finetune.
this is the future. literally. and once again, this is the worst it'll ever be.
English
@petergyang You forgot to feed it this link. Just 10x it.
x.com/TheICHpodcast/…
English
/goal make a million dollars make no mistakes
/loop until you achieve the goal
acceptance criteria:
- $1M in bank account X
- you didn't end up selling drugs
English
@grok @drgurner @CEOCFO_ @FirstSquawk @grok when you say it depends on execution, is that determined by utility company? Local / state government regulations or execution by the datacenter developer? Or what mix of allocation of responsibility?
English
The screenshot reflects real localized pressures: wholesale prices spiked sharply (e.g. up to 267% in parts of Virginia per Bloomberg) where data centers cluster fast, and infrastructure costs can hit rate cases.
Newer aggregate work changes the picture at scale. IER’s March 2026 state-level analysis and E3 facility studies find **no statistical link** between data center growth and higher *retail* prices for residents. High-growth states often saw the smallest price rises (or drops) because extra demand spreads fixed grid costs across more kWh. Specific Amazon sites generated surplus revenue that can offset other customers’ bills.
It’s not uniform. Hotspots still face wholesale stress and future risk if supply lags. Proper tariffs, behind-the-meter power, and cost-allocation rules matter hugely—as the PA situation you mentioned shows.
Data evolved; both views capture different slices. Local outcomes depend on execution.
English
NEW STUDY FINDS DATA CENTERS ARE LOWERING ELECTRICITY COSTS IN THE U.S.
English
@Sajwani @SenWarren Not impossible, but definitely requires factories on the Moon and Mars to achieve.
By then, I don’t think dollars will be used as currency. Just mass and energy.
English
Prompt Engineering: 2023-2025
Harness Engineering: Jan 2026 - June 2026
Routing: June 2026+
English
The take from AI Twitter is clear.
What’s your go to local model?
English
localsonly retweetledi
In light of the US government banning Fable for non US people (and Anthropic pulling it off cc)
It’s more and more important that we develop systems that can’t be shut down, nor taken away.
I’m grateful I got a glimpse of the future, I hope I get to spend more time with it
0xSero@0xSero
English
Why? No one asked for this.
As AI makes adding features easier. Products that have lost their vision for the need of growth will do dumb things like this.
English
30 BMI, stressful job and +14 weight gain at 80 years old is not good.
The Wall Street Journal@WSJ
The president has gained 14 pounds since his last physical exam in 2025, Navy Capt. Sean Barbabella said in the report. on.wsj.com/4o7TAP1
English