StevenFu

3.7K posts

StevenFu banner
StevenFu

StevenFu

@StevenFu

Most my comments are Sci-Fi plots, don’t take them seriously please.

Palo Alto, CA Katılım Haziran 2009
321 Takip Edilen159 Takipçiler
Peter H. Diamandis, MD
Peter H. Diamandis, MD@PeterDiamandis·
.@elonmusk just crossed $800 billion — roughly 2.7% of the entire US GDP. The last person to hold that much of the American economy? John D. Rockefeller in 1913. It took a century for anyone to match him. Rockefeller had oil. Musk has the future.
English
466
642
13K
1.1M
⭕️ CyberMike ⭕️
⭕️ CyberMike ⭕️@CyberMikeOG·
Who really is going to take the time to buy any of this when it’s all locked up. I just walk out of the store. A typical store will spend $800k on these locked cabinets. They won’t pay back the losses.
⭕️ CyberMike ⭕️ tweet media
English
11
0
18
645
StevenFu
StevenFu@StevenFu·
@CyberMikeOG She should lose freedom for life, that’s clearly attempted murder.
English
0
0
1
6
Irushi
Irushi@Im_IrushiK·
Guess the legend.
Irushi tweet media
English
479
603
16.9K
4.5M
StevenFu
StevenFu@StevenFu·
@stevenmarkryan It is slow AF though, practically trapped us in a single solar system. Well…our wetware at least.
English
0
0
0
30
stevenmarkryan
stevenmarkryan@stevenmarkryan·
Do you ever think about how insane the internet is? (and how lucky we are that the speed of light ain't slow AF)
English
19
9
212
9.1K
Shibetoshi Nakamoto
Shibetoshi Nakamoto@BillyM2k·
why do i have 2.8 million followers now if you’re not a bot say hi so i can see my real number
English
2.6K
76
2.8K
153.4K
StevenFu
StevenFu@StevenFu·
@realpeteyb123 These are going to confuse archeologists for generations to come.
English
0
0
0
1
Peter B
Peter B@realpeteyb123·
Took the triplets to a desolate beach for their birthday. Had them build rock people. Promised them we will go to Internet to choose who made the best rock person. Drop a comment: Kid A Kid B Kid C Winner gets to plan itinerary tomorrow.
Peter B tweet mediaPeter B tweet mediaPeter B tweet media
English
2.8K
113
2.3K
212.8K
StevenFu
StevenFu@StevenFu·
@Tsla99T Unsupervised 必须涨价,不然不math
中文
0
0
0
227
StevenFu
StevenFu@StevenFu·
Opinion: We have already hit physical limits on DRAM, NAND, and magnetic disks. Unless we have some breakthrough in other types of storage, it would be hard for unit prices of these to fall again. Plenty of new fabs are needed.
English
0
0
0
15
StevenFu
StevenFu@StevenFu·
@wholemars These are often just old people who were convinced by MSM that if they don’t have a sticker like that, their car will be keyed all day every day.
English
0
0
0
46
Whole Mars Catalog
Whole Mars Catalog@wholemars·
I actually love these bumper stickers. A lot of Tesla fans get mad, but I just laugh. People love their Teslas so much that they’d rather buy a virtue signaling bumper sticker than switch to a different brand.
Gail Alfar@gailalfaratx

Just spotted a Model Y with a sticker on it: “Anti-Elon Tesla Club.” Stop advertising your stupidity. If your kid doesn’t know how to swim and you’re wearing a life vest made by “Javier,” would you plaster a label on that life vest that says “Anti-Javier Life Vest Club!”?

English
35
23
457
32.9K
StevenFu
StevenFu@StevenFu·
Professionals using them that way doesn’t necessarily mean that’s the right way. Many professionals were exposed to models before the curve and have mental scars. More often they are constrained by the company to only sub-par models, which they’re right to be skeptical of. This doesn’t represent what SOTA models can do.
English
0
0
0
79
Sukh Sroay
Sukh Sroay@sukh_saroy·
A new study just blew up the entire "vibe coding" movement. Researchers from UC San Diego and Cornell tracked 112 experienced software developers using AI agents in their actual jobs. The finding is the opposite of every viral demo on your timeline. Professional developers don't vibe code. They control. Here's what they actually found. The researchers ran two studies. 13 developers were observed live as they coded with agents in real production work. 99 more answered a deep qualitative survey. Every participant had at least 3 years of professional experience. Some had 25. The viral pitch of agentic coding goes like this. Hand the agent a vague prompt. Don't read the diff. Forget the code even exists. Trust the vibes. Andrej Karpathy coined the term. Tens of thousands of developers on X claim to run "dozens of agents at once" building entire production systems hands-off. The data says almost nobody serious actually works that way. Here is what experienced developers do instead. → They plan before they prompt. They write out the architecture, the constraints, and the edge cases first, then hand the agent a tightly scoped task. → They review every diff. Not because they're paranoid. Because they've seen what happens when you don't. → They constrain the agent's blast radius. Small, well-defined tasks only. The moment a problem touches multiple systems or has unclear requirements, they take over. → They treat the agent like a fast junior dev that needs supervision, not a senior engineer that can be trusted alone. The researchers also found something darker buried in the data. A separate randomized trial they cite showed that experienced open source maintainers were 19% slower when allowed to use AI. A different agentic system deployed in a real issue tracker had only 8% of its invocations result in a merged pull request. 92% failure rate in production. 19% productivity drop for senior devs. The viral demos lied to you. The paper's biggest insight is in one sentence: experienced developers feel positive about AI agents only when they remain in control. The moment they let go, quality collapses, and they know it. This matches what every serious shop has quietly figured out. The developers shipping the most with AI right now aren't the ones vibing. They're the ones with the strictest review processes, the tightest task scoping, and the clearest mental model of what the agent can and cannot do. Vibe coding makes for great Twitter videos. It does not make great software. The next time someone tells you they let Claude build their entire SaaS in a weekend, ask them how much of that code they've actually read. The honest answer separates real engineers from the demo crowd.
Sukh Sroay tweet media
English
196
331
1.7K
259.5K
grant
grant@grantbelden·
I will chop my dick off on livestream if Tesla exceeds Waymo in 6 months
Gene@genejchan

@grantbelden Wow that means Waymo is in deep trouble That's only 7-8 doubles from where $TSLA is, which means Robotaxi will exceed Waymo in about 6 months

English
85
20
277
79.3K
StevenFu
StevenFu@StevenFu·
@RedWavePress You can buy a used Tesla model 3 for less than Corrola money, guess most people don’t even know.
English
0
0
1
75
RedWave Press
RedWave Press@RedWavePress·
SHOCKING: The U.S. Department of Agriculture is working to close a loophole that allows individuals to access taxpayer-funded food stamp benefits without states verifying their physical and monetary assets. This gap has enabled numerous recipients who own LUXURY vehicles to continue receiving benefits. Fox News: “The Department of Agriculture tells us in one red state alone SNAP participants own more than 14,000 luxury cars that include 3 Ferraris, 11 Lamborghinis, 59 Maseratis, 141 Porches, and more than 2,000 Teslas.” “Right now 42 states have the program.”
English
47
618
2.2K
59.5K
StevenFu
StevenFu@StevenFu·
Wait what? How could they lose 5B a year? Is it just negative cash flow of 5B and most ends up as assets? Or they really manages to lose 5B?
Gary Black@garyblack00

@wholemars WS expects $GOOG to lose $5B in 2026. At GOOG’s targeted 1M Waymo paid rides per week, that’s about $100 loss per ride (-$5 billion / 52 million). GOOG has previously targeted 2027 year-end for Waymo to turn profitable.

English
0
0
0
13
StevenFu
StevenFu@StevenFu·
@mpgenchau @wholemars Those are literally 3D moves. They prefer to stay on the Earth's 2D surface, and it would be even better if it were flat.
English
0
0
1
15
Michael Phan
Michael Phan@mpgenchau·
@wholemars Some people can't see 4D chess moves. Those are the means to the end that he desires.
English
1
0
0
33