Paata Matikashvili

[Paata Matikashvili @PaataMatika]

My grandpa always knows exactly what to say
Canceled plans to do cycling instead, no regrets

[🌺Sai🌺 @Sai60012]

Teri Mahima aprampar Khatu wale Shyam🚩
K
h
a
t
u
w
a
l
e
S
h
y
a
m
S
Shish ke Dani Maha balwani 🚩
h
i
s
h
k
e
D
a
n
i
M
a
h
a
b
a
l
v
a
n
i
K
h
a
t
u
w
a
l
e
S
h
y
a
m
S
h
i
s
h
k
e
D
a
n
i
M
a
h
a
b
a
l
v
a
n
i
Sabhi per apni kripa Drishti banae rakhen Shyam 🙌 @grok

[Paata Matikashvili @PaataMatika]

@prudentmoose that kind of quiet safety hits different 💛

[Paata Matikashvili @PaataMatika]

If you're the smartest person in the room you're in the wrong room
My dog fell asleep in the most ridiculous position and they looked so proud of themselves

[Paata Matikashvili @PaataMatika]

@jose41814833 That's facts. The comeback always comes stronger 💪

[Paata Matikashvili @PaataMatika]

Instant serotonin from the smell of the ocean
Overthinking is just creative problem-solving without the solution

[Paata Matikashvili @PaataMatika]

Instant serotonin from the smell of the ocean
Overthinking is just creative problem-solving without the solution

[Paata Matikashvili @PaataMatika]

The world would be better with more pandas
If you haven't tried Italian pasta you're missing out

[Overclock @nverbullish]

A computer store is now teaching customers how to stop paying for ChatGPT. In a 60-second video. For free.

Micro Center. The same place that sells you the GPU is now showing you how to replace the subscription it powers.

The tool is called LM Studio. Free. One download. No terminal. No command line. A clean GUI that lets you browse, download, and run local LLMs like you're scrolling an app store.

Llama 3. Phi 3. Gemma. Dozens of open-source models sitting in a library. Click one. It downloads. Click run. It answers.

Meta-Llama-3-8B-Instruct. 4.92 gigs at Q4 quantization. Fits on a laptop with 16 gigs of RAM. Full GPU offload. The model runs entirely on your graphics card.

He asks it how to run AI locally using LM Studio. It explains itself. Step by step. Running on 63 gigs of RAM and an 8-gig GPU. No internet. No API key. No account.

ChatGPT Plus: $20 a month. $240 a year. Every prompt logged on someone else's server.

LM Studio: free. The model: free. The weights: a download link from Hugging Face. The conversation: stays on your hard drive.

A hardware retailer just told its own customers they don't need the cloud anymore. That's not marketing. That's a signal.

When the store selling the parts starts teaching you how to leave the subscription, the subscription is already dead.

The checkout line at Micro Center is the off-ramp from the cloud.

[Paata Matikashvili @PaataMatika]

@jose41814833 nothing hits quite like that old book smell

[Paata Matikashvili @PaataMatika]

Nobody warns you that 27 comes with an existential crisis - Thục Đ. Liễu

[GENG @g164843639]

The hardest part is starting. - Đây K. Liễu

[Paata Matikashvili @PaataMatika]

The peace I find in golf is unmatched. - Cầu T. Phạm

[Paata Matikashvili @PaataMatika]

The best part of visiting Seoul was the local bossa nova scene. - Chơn P. Bùi

[Prophet @prophetmarketai]

More than 3,000 markets are now open on Prophet! Browse existing markets or create your own.

Test your forecast against The AI Machine with instant liquidity: https://t.co/fbfSYkrq9x

No US users allowed.

[Paata Matikashvili @PaataMatika]

@pabnaindia Exactly

[PABNA @pabnaindia]

दिल्ली-देहरादून एक्सप्रेसवे हादसे से लें सीख: सड़क पर भूलकर भी न करें ये 5 गलतियां, जा सकती है जान
https://t.co/WW6LMtlhKF
#ऑटो

[Paata Matikashvili @PaataMatika]

Just reviewing documents is somehow both boring and stressful - Thủy H. Thủy

[Flandermaxx @Flandermaxx]

A 28 YEAR OLD KRAKOW MAKER PAIRS JETSON ORIN NANO BOARDS WITH USED RTX 3050s, SHIPS HANDHELD AI GAMING RIGS FOR $1,899, PULLED $30,384 LAST MONTH

bartek is 28, krakow flat with a 3d printer humming on a flowery linen table runner, sources jetson orin nano dev kits at $499 each through a polish edu channel

pairs each with a refurbed RTX 3050 6GB at $120 from a gielda komputerowa in nowy targ, 3d prints a steamdeck style shell, the rig boots in 4 seconds

pause at 0:30 on the silver heatsink wired into the rtx, that is one box running cyberpunk at 1080p while a local llama 3.2 3B NPC talks through the headset

16 indie game devs bought one last month at $1,899 each to prototype AI NPCs in their unity builds, $30,384 in

BOM $782 per unit, margin $1,117 per rig, the first sale of the month covers parts for 8 more

the window is open, follow and bookmark before it closes

[Eileen's mother ❣🍉 @rola_almoghrabi]

I NEED DIAPERS😭😭

︎
︎

︎

︎

︎
︎

︎

︎

︎

we’ll remind you 🥺
︎

︎

︎

︎

︎
︎

︎

︎

︎

︎
︎

︎
︎

︎
︎

︎

︎

︎

︎
︎

︎

︎

︎

that
︎

︎

︎

︎

︎

︎
︎

︎

︎

︎
︎
💔
︎

I desperately need diapers for my baby 👶🏻
https://t.co/FLk3iM7URh