Zachary Levy (@ZacharySunforge) header
Zachary Levy (@ZacharySunforge) profile picture

Understand Pepega, I would use the Ring from a desire to do good. But through me, it would wield a power too great and terrible to imagine.

Nevada
Joined October 2023
Photo by Zachary Levy (@ZacharySunforge)Photo by Zachary Levy (@ZacharySunforge)Photo by Zachary Levy (@ZacharySunforge)Photo by Zachary Levy (@ZacharySunforge)Photo by Zachary Levy (@ZacharySunforge)Photo by Zachary Levy (@ZacharySunforge)Photo by Zachary Levy (@ZacharySunforge)Photo by Zachary Levy (@ZacharySunforge)Photo by Zachary Levy (@ZacharySunforge)Photo by Zachary Levy (@ZacharySunforge)
The climate agenda is a deadly socialist power grab.
2
5
1,254
Zachary Levy retweeted
Imagine following nearly 13 million real people for more than 10 years just to see what cholesterol level helps them live the longest. That is exactly what Korean scientists did in a giant 2019 study. They tracked 12.8 million adults who had check-ups between 2001 and 2004, and they kept watching until 2013. More than 694,000 people died during that long stretch. The results were clear and surprising. Cholesterol and death formed a U-shape. Very low cholesterol was linked to higher death rates. So was very high cholesterol. The sweet spot for the longest life was 210 to 249 mg/dL for most adults. That is higher than many doctors call “ideal.” Younger people had a slightly lower best range, but still not the super-low numbers often pushed today. Because the study lasted so many years, reverse causality is less likely to explain the low-cholesterol danger. Reverse causality means a hidden illness might lower your cholesterol right before it kills you. With more than a decade of follow-up, early sickness cannot easily account for the extra deaths seen in people who started with low cholesterol. The pattern held across ages and for both men and women. The huge number of people—almost the entire adult population of a country—makes the findings hard to ignore. Smaller studies can be wrong by chance. This one is massive. It shows that moderately higher total cholesterol, not the lowest possible number, went with the best chance of living longer in this large group. In short, the biggest study of its kind found that cholesterol levels in the low-to-mid 200s were tied to the longest lives, and the long tracking time makes it tougher to blame the higher death rates at low levels on reverse causality alone.
197
1,285
95
5,372
392,415
Zachary Levy retweeted
This doesn't mean it's not gonna get better. This doesn't mean AI is useless. The technology is amazing, you can do great things with it. But you need to actually care and work as hard as you ever did. People who think that you don't need to get good anymore are ngmi
>see "Adobe is done" all over timeline >wow, entire suit rebuilt >"Every Crafting App is written from scratch in Rust and held to the same rules." >"native, not wrapped" >Adobe must actually be over >download one app >5gb RAM idle, 100% CPU, not responding >look inside >Tauri They said "pls no electron or tauri", AI did it anyway, but sloppers don't look at the code or use their own stuff. I think Adobe will be fine
Readers added context they thought people might want to know
The Crafting Apps shown (e.g. PhotoCraft) use native egui/eframe, not Tauri. The separate ArtCraft AI app uses Tauri. Poster confirmed checking ArtCraft. github.com/storytold/phot… getartcraft.com/apps
14
11
147
4,828
Zachary Levy retweeted
Executive statements on this are 100% predictable based on whether or not they will profit from government subsidization / bailouts / regulatory capture. There is no threat whatsoever. The centralized model services want regulatory capture because that is their only sustainable path in the market; if the market progresses without distortion, local models on local hardware win. This is a win for Nvidia. It makes the already unsustainable economics for centralized model services even less sustainable. This is why Sam Altman says what he does, it’s why Dario says what he does, and it’s why Jensen says what he does.
JUST IN: Nvidia CEO Jensen Huang declares AI doomer predictions are “made up.”
8
27
4
440
15,177
Breaking: Michael Burry says OpenAI and Anthropic are faking the AI warnings due to slow growth ahead of IPOs Below are his reasonings: 1. LLMs are not AI and won't be AGI. There is nothing AI to slow down. 2. Competition is coming up fast, slowing benefits incumbents. 3. IPOs need hype & puffery; "we are so awesome it could become dangerous" is hype & puffery 4. Cover for real uncontrollable slowing growth as IPOs look to be pushed out
607
3,191
482
20,225
1,758,919
The Anthropic and OpenAI bailout and regulatory protection is gonna a fuckin bitter pill to swallow but I can't imagine any scenario where it doesn't happen
3
93
Zachary Levy retweeted
Replying to @ThePrimeagen
In my view, their fearmongering has less to do with selling to the public, and much more to do with selling to investors and government. If they condition the public to believe their technology is extremely dangerous, then it will be politically viable to build a regulatory moat.
18
19
632
16,400
Zachary Levy retweeted
European tech is simply unstoppable. It's over for Silicon Valley
197
824
153
8,519
709,691
Zachary Levy retweeted
110
2,542
114
30,046
664,725
It's not surprising to see handmade software come out more impressive than billion-dollar incumbents. They've enjoyed decades competing against nobody, and made a monster of bureaucratic mediocrity with the excuse of "the cost of complexity." The bar is miles underground.
2
19
261
5,291
Zachary Levy retweeted
Me explaining AI to friends
577
4,761
966
29,951
5,819,257
Zachary Levy retweeted
"The N64 shipped with the NEC VR4300: a single core, in-order execution, 93.75 MHz CPU. It was paired with the 'Reality Co-Processor' (RCP), running at 62.5 MHz, for acceleration of audio and graphics computations. It supported 4 MB of RAM (later expanded to 8 MB with the Expansion Pak). Game cartridges for the N64 were, at most, 64 MB. "By today’s standards, this is incomprehensibly underpowered. You, the reader, in your pocket—or perhaps in your hand, while you read this—likely have at least four out-of-order execution cores, running somewhere from 2 GHz to 4 GHz. That’s something like 20-40 times more cycles per second than the N64’s VR4300, spread across 2-4 more cores. Each core is vastly more capable of higher throughput in doing useful work, given their out-of-order execution, deeper pipelines, new caching layers, and numerous other hardware optimizations that I’m not qualified to even attempt to explain. This magical device in your pocket also has 8 GB of RAM—three orders of magnitude more than what the N64 offered. It also can have something like 32, 64, 128, 256, 1024 GB of cold storage—several thousands of times larger than the N64 cartridge size. "Needless to say, computing technology has progressed in countless ways since the 1996 release of the N64. But the belief in the constant of progress over time is a privilege unique to modern civilization, and this belief can blind to concrete examples of regress over time. "Despite the aforementioned mind-boggling computing hardware advancements, there remains enormous love for the N64, and other old computing devices. The Analogue 3D, an FPGA implementation of original N64 hardware, released in 2025. Similarly, the ModRetro M64 (another such implementation) released in 2026."
3
23
2
147
24,430
If @OpenAI and @AnthropicAI don't get the state to build them a significant regulatory moat in the next couple years they are completely fucked. Shitload of resources are going to be poured into fighting tooth and nail for this.
2
46
I want to do some vibe coding but all the vibe coding tools are vibe coded and thus do not work. Is there a name for this vibe coding paradox?
1
42
thinking about this post again
112
7,224
128
141,913
1,572,347
I think you should do something else than design.
Celebrating small wins is hardwired into "Achieve," our goal-tracking sample app. We wanted the simple act of checking off a task to feel like a celebration. Instead of the standard #Material ripple, we used Styles to drop in a custom, high-contrast completion animation—specifically, a brand-colored ring that expands and fades on tap, paired with a subtle pressed-state gradient. It lets us focus entirely on polishing those custom, rewarding micro-interactions while Material handles the foundation. #GoogleDesign #MaterialDesign #JetpackCompose #AndroidDev #UXDesign #InteractionDesign
4
32
2,475
Zachary Levy retweeted
Replying to @sudoingX
2
10
1
110
20,120
Zachary Levy retweeted
Intellect should be judged not by the words a person say but by the outcome of their actions. Not even the actions themselves but their outcome. Chess grandmasters make seemingly stupid moves until you suddenly realize that you lost.
48
95
5
1,577
49,278
Zachary Levy retweeted
As you become an adult, you realize that things around you weren't just always there; people made them happen. But only recently have I started to internalize how much tenacity *everything* requires. That hotel, that park, that railway. The world is a museum of passion projects.
670
11,665
1,452
63,948
Zachary Levy retweeted
US AI labs mad about China distilling their models
23
246
12
2,254
73,364
Zachary Levy retweeted
When AI companies pirate something*, it's "fair use". When someone pirates their model data, it's "stealing" and "undermining American research". * - Literally pirate. Multiple court admissions have made clear that AI companies download millions of works from pirate sites.
We have information that Moonshot AI distilled Anthropic’s Fable for the development of its K3 model. To do this they developed a sophisticated internal platform to conduct large scale distillation against U.S. models, allowing them to quickly switch between multiple methods of access to avoid detection. Moonshot AI has also acquired GB300-equipped servers and has accessed GB300s in Thailand, likely to train its AI models.   The United States strongly supports the free and fair development of AI, including a thriving competitive ecosystem that spans frontier models, specialized systems, open-source frameworks, and open-weight models. Legitimate AI distillation used to create smaller, more efficient models plays a vital role in this open innovation ecosystem. However, large-scale, covert industrial distillation aimed at stealing proprietary U.S. technology and undermining American research is unacceptable.
52
224
7
2,465
68,343