It's over an order of magnitude heavier than the largest of those (Helios).

There's a difference between an aircraft being built to lift itself and perhaps a tiny payload and be paperlight vs. an aircraft designed for actual commercial service carrying passengers.

The feet are needed to accelerate the plane to takeoff velocities in order to conserve battery mass.

Also, the big AI companies get a number of advantages, such as large batch sizes and being able to balance on-peak and off-peak workloads better.

Is most of your daily driving "a race"?

This is a midrange model. This is competitive midrange pricing. For my midrange work I've been using GLM-5.2 - Gemini 3.7 flash is a bit cheaper and a bit better on the benchmarks, so yeah, I'll probably switch so long as the prices stay low.

Sometimes you need a high-end model (Claude Fable, GPT 5.6 Sol, Kimi K3, etc). Sometimes a low-end model is fine (GPT 5.6 Luna, DeepSeek 4 Flash 0731, etc). Usually a midrange model is the best balance.

If the original ToS on Twitch grants Amazon a license to create derivative works then they don't have anything to worry about.

It's like your local superstation had the rights to run ads on Star Trek in addition to the license fee they paid but they didn't therefore get a right to create a new Star Trek series or publish a "Best Moments of STTNG 1989" show around the holidays.

Now if they had gone to Paramount about the Best-Of show in theory they could have negotiated a license for that, having already been media partners. Heck, some software company got Riker to pitch their product from the Bridge.

The big authors' settlement with OpenAI sets at least some precedent in this nascent field of IP law (which is enforced whether we like it or not).

But I've never heard of a Twitch streamer talking about a derivative work that Amazon created from their show using their words or likeness.

To be fair Youtube does indeed do a yearly "best of" show.

Who knows, maybe we'll see an AI Hassan Piker doing and saying some wildly ill-advised shit in the near future. Would not be entirety shocking.

There are to be three variants of the ES-300: an all-electric 200km version, a hybrid 30-seat 400km version, and a hybrid 25-seat 800km version. And the ES-300 airframe very much is not some existing airframe.

EV fires are rarer than ICE fires (and slower spreading as well), so given that the competition is fuel-based....

It is a perfectly cromulent statement. 1MW = power, $5 = a cost of energy. They are providing both power and (indirectly) energy.

They didn't bother to install a co-pilot, a seat for the co-pilot, passengers, seats for the passengers, the hybrid gen set as you say, fuel for the gen set, and who knows if they installed a full production sized battery or just a smaller one

Are you just learning about the concept of a prototype in this comments section?

That's not correct. There are three modes. All-electric, hybrid-30 passenger, and hybrid-25 passenger, with ranges of 200km, 400km, and 800km, respectively. The airframe is to be certified for 20k feet (FL200), with optimal cruise altitudes from 10-15k feet, but this was explicitly described by the company as a "low-altitude test flight".

And to all the people going, "LOL, that's never going to cross the Pacific!" or similar range comments... and? Long haul flights (over 4000km) are only 5% of commercial flights. Very short / regional flights are 27% of total flights. There's very much a market for puddle jumpers, and airlines very much will buy them if they offer competitive advantages.

This is smart but harder than it needs to be.

Do we have a wrapper available that can let native Zoom launch and only have access to ~/.config/zoom and some USB input devices but not the rest of the machine?

Somehow building a custom MAC profile is expert-level stuff two decades on and not an end user's job.

An on-demand screen sharing permission popup should be achievable.

It's silly that a browser sandbox is better than any native sandbox available.

Perhaps such a tool readily exists and just needs to be mainstreamed into existing distros.

I know there was a Debian project a while back that aimed for all services to pass systemd strict security by default but nobody wanted to do the work.

That's the right mindset but separate from user apps.

There's some interesting stuff going on trying to understand more of the atomic structure beyond the current models.

The math on gluons has worked quite well but for some reason questions like, "OK, but what are gluons?" has been derided as "crank science" in Pergamon Press-gated Academia.

Such a diversion from Science from the beginning of Human history through the 1960's.

Structures like DeLay-Bendebury Fibers and other models of what's going on behind the math are happening outside of Academia where Physics PhD's are discarded and PhysEd BA's are celebrated.

Whichever model comes next will likely happen adjacent to the podcaster sphere and be reduced to useful engineering before Academics take a look at it.

There are plenty of good researchers working in institutions but for fundamental revisions at this point they all seem to need to leave Academia and go out on their own seeking private investment.

Which is fine except that we're all paying for the uninterested remainder. The great advancements that make our lives better always begin in fundamental science generating new models that better describe this reality. We don't get much from a jobs program for Mathematicians playing Physicists and it's very expensive.

It's quite strange that most people find the status quo completely acceptable and are willing to throw another $100B at a higher energy collider but are unwilling to grant base-layer investigations that are nearly free in comparison.

If we believe the former head of Skunkworks they've recruited the best and brightest since the 1970's and had made breakthroughs by the early 90's in at least topological physics. So we get to pay for that but aren't allowed to use it which is intolerable. But it's also understanble in a pre-Internet era where the only feasible choices were String Theory or black projects.

1) 30B, dense. Optimized to fit Q4 quantized in a 24GB card with a speculative decoding model as well.

2) Because reporters don't know what weights are and assume you don't know either.

3) Way better than Gemma 4 on text tasks, slightly better on multimodal. Numbers below are all: Benchmark: Muse Glimmer score Gemma 4 31B score difference

Artificial Analysis Intelligence Index: Muse Glimmer: 35 30 +5
MCP Atlas (Public): 75.5 54.2 +21.3
DeepSearch QA: 74.6 61.7 +12.9
SWE-Bench Pro: 51.2 36.9 +14.3
SWE-Bench Verified: 76.0 66.6 +9.4
OSWorld-Verified: 65.9 58.5 +7.4
GAIA2: 43.3 36.4 +6.9
WildClawBench: 47.6 37.6 +10.0
TerminalBench 2.1: 51.7 43.4 +8.3
Tau3-Banking: 23.5 15.1 +8.4
MMMU Pro: 74.0 73.0 +1.0
Charxiv Reasoning: 78.8 77.7 +1.1
OmniDocBench v1.5: 75.8 72.5 +3.3
ScreenSpot Pro: 75.4 75.9 -0.4

4) 128k tokens

5) No, sadly.

What you want is a DGX Spark.

It costs $4000.

And yes, there are fundamental advantages to cloud services, such as large-scale batching, little idle downtime, high speed, and hardware optimized to the specific models / serving needs. That said, one can weigh that off against sovereign control over your server...

Define "tolerable".

The model in question - Muse Glimmer - with DFlash/speculative decoding - will probably get you ~60 to 124 tok/s on a 3090 (a quite dated GPU). For a 5090, it's said to clock in at 233,4 tok/s. On CPU you're looking at maybe 3-5 tok/s.

If you call that "tolerable", I guess you're more patient than me? And as mentioned, you're not just wasting time, but also wasting a lot of power too - CPU is a very power-inefficient way to run ML models.

If you insist on CPU, this isn't the right kind of model anyway. You want to take advantage of the fact that you probably have lots of (comparably cheap) RAM, and compensate for the fact that you have (comparably) terrible memory bandwidth, and for that, you want a MoE with a high total parameters but a low active parameters. Not a dense model like this.

Then I was basically aiming at Mac Minis

That's very much a special case which you didn't mention in your post that I responded to, but still the answer is "meh". You couldn't run it at all on a 16GB Mac Mini, and I think you'd struggle to run it at all on a 24GB (because you have to share the ram with the OS, the inference server, etc). For the base Mini you might get 10-12 tok/s, and for the M2 Pro / M4 Pro, maybe 25-35 tok/s. Still pretty far from a GPU, though.

This model is designed for >= 24GB GPUs.