Fabio Biola

Ron Weasley yelling “Expensive Petroleum” while filling up his car might be the peak of comedy 🤣

@justbyte_ Less is more.

Share your desktop wallpaper I'll start:

@LensScientific Because that's how we are wired. Time doesn't exist, we make time. You can check the long explanation in "The order of time" from Carlo Rovelli.

If the laws of physics don’t prefer a direction of time, why does everything we experience act like time has a preferred direction?

@poetengineer__ Love the concept! Do you think you would share it on GitHub at some point? I would love to test it on my projects

visualizing all my claude code sessions: each plant is a project, its leaves the sessions. sliding the timeline scrubber -> replaying projects growth over time.

@mathemetica So, can you tell me about the camera you used to get this view?

This isn’t CGI. This is real mathematics controlling atoms inside every processor on Earth. Watch until your brain restarts.

@johanvinet I appreciate you take time to do anything by hand, but man isn't there any kind of tool that can help you? (I'm not talking about AI) Even if I'm sure you like to do it I can't imagine all this hard work to always be rewarding by itself. But maybe that's just me

Also, LUNARK is human-made, frame by frame. No AI involved, just pure dedication and love for the craft. 🥰

@RickNas95547619 @SaqibBanbh90290 @ItsEric208 I guess it depends if you want to give instructions to your tiler or responding to a lateral thinking quiz

@NeoKree @SaqibBanbh90290 @ItsEric208 Using this logic...the linoleum on the floor has a multitude of squares. They must also count.

Number of saqure you see...? ONLY 1 % Will succeed

@SaqibBanbh90290 @ItsEric208 I can guess: 3x3 squares. 1 square containing all the squares 1 square for the biscuit 1 square for the image post 1 square for the website containing the image post 1 square for the screen containing the website the question is: why does your phone/laptop have a square screen?

@ItsEric208 How Explain

@thetransportcr Does this effect have a name?

took the advice to take photos with this "filter" before my camera dies

@Rainmaker1973 Until you are with your friends and people really do it with 6-7 meters below it

What you see on social media can sometimes be a bit deceiving.

@lassvestergaard simple fix, don't use Opus on Claude Pro. If you stick to Sonnet you can code for 3-4 hours before hitting limits. Which I think it was very intentional (time window is 5h). But still getting things done.

never buy claude pro, this is after using opus 4.6 with plan mode 2 times and just redo a landing page 2 times. All took max 4-6 minutes 🥲

@tchaloyi Do you also know how Apple solved this issue?

Back in 2011, I went into computer hardware repairs. I started with laptop chargers. I remember peeling back the rubber, soldering wires, and always staring at that weird plastic bulge on the cable. It didn't look like it did anything, but it was on every single high-end charger I fixed. I used to wonder if it was a hidden battery or just a weight to keep the cord from tangling. It turns out, that little lump is the unsung hero of your workspace. It's called a Ferrite Bead, and its only job is to act as a silencer for your electricity. See, every electronic device is naturally noisy. They send out invisible electromagnetic signals. Without that cylinder, your charger cable would turn into a giant antenna, broadcasting interference that would make your Wi-Fi slow, your TV flicker, or your speakers buzz. Inside that plastic shell is just a chunk of magnetic iron. It catches all that electrical noise and kills it before it can escape the wire. It’s basically a muzzle for your cable so your gadgets can live in peace. INALEGWU.

@LiorOnAI @liquidai Can you release official agentic benchmarks?

A 24-billion-parameter model just ran on a laptop and picked the right tool in under half a second. The real story is that tool-calling agents finally became fast enough to feel like software. Liquid built LFM2-24B-A2B using a hybrid architecture that mixes convolution blocks with grouped query attention in a 1:3 ratio. Only 2.3 billion parameters activate per token, even though the full model holds 24 billion. That sparse activation pattern is why it fits in 14.5 GB of memory and dispatches tools in 385 milliseconds on an M4 Max. The architecture was designed through hardware-in-the-loop search, meaning they optimized the model structure by testing it directly on the chips it would run on. No cloud translation layer. No API roundtrip. The model, the tools, and your data stay on the machine. This unlocks three things that were impractical before: 1. Regulated industries can run agents on employee laptops without data leaving the device. 2. Developers can prototype multi-tool workflows without managing API keys or rate limits. 3. Security teams get full audit trails without vendor subprocessors in the loop. The model hit 80% accuracy on single-step tool selection across 67 tools spanning 13 MCP servers. If this performance holds at scale, two assumptions need updating. First, on-device agents are no longer a battery-life trade-off; they're a compliance feature. Second, the bottleneck in agentic workflows is shifting from model capability to tool ecosystem maturity.

@ramin_m_h @liquidai Nice but we want benchmarks, not speed inference. Can you do the same ones Qwen did?

Truly impressive release of hybrid tiny models from the Qwen team as always! People are asking how do they compare in speed, latency, and memory to @liquidai’s LFMs for on-device deployment? Here is a quick profiling on Apple M3 Ultra: > LFM2.5-1.2B is 52% faster in decode than the Qwen3.5-0.8B. > LFM2-700M is 71% faster than Qwen3.5-0.8B on decode > LFM2-2.6B has the same speed as Qwen3.5-2B on decode > LFM2-700M uses 46% less peak memory than Qwen3.5-0.8B > LFM2-2.6B uses 21% less peak memory than Qwen3.5-2B > lfms prefill with the same parameter size is generally 12% faster than Qwen3.5 We designed LFM2 series with our hardware-in-the-loop meta AI design approach that allows us to find out the most efficient architecture for a given processor without quality sacrifice. This test is done on Apple M3 Ultra, 512 GB unified memory Config: > 512 prompt tokens, 128 generation tokens, > 5 trials per configuration > Framework: MLX (mlx-lm / mlx-vlm)

@rohanpaul_ai @grok is this image true? Please cite sources if you know them

Big revelation in this data. Everyone keeps talking about an AI bubble, but we forget the fact that only 0.3% of the global population actually pays for a premium subscription. For the vast majority of the real world it has not started yet. 6.8B people have yet to interact with even 1 free chatbot. The massive grey chunk is the 6.8B people who have zero experience with AI. The green dots represent 1.3B people using free versions of tools. Only a small group of 15-25 mn people pays for subscriptions. We sit in that microscopic slice.

I just released my first AI trailer: OFF GRID youtu.be/u2phtkmtLy4 Let me know what you think! #aifilm #indiefilm #aivideos

@maximelabonne Are there official benchmarks against gpt-oss and Qwen3?

Early checkpoint of our biggest LFM2 model to date 🎉 It shows good scaling and extremely fast inference vs. gpt-oss-20b and Qwen3-30B-A3B We'll release an LFM2.5 version with more pre-training and RL in a few months

@wildmindai What's the commercial price for the board?

17,000 tokens per second!! Read that again! LLM is hard-wired directly into silicon. no HBM, no liquid cooling, just raw specialized hardware. 10x faster and 20x cheaper than a B200. the "waiting for the LLM to think" era is dead. Code generates at the speed of human thought. Transition from brute-force GPU clusters to actual AI appliances. taalas.com/the-path-to-ub…

@Hesamation Him moving and acting like a crab 🦀 is perfect

YoU’rE gOinG tO bE LEFt beHind! 🤬 - the average OpenClaw user so accurate 😂