Vinay Y S

[Vinay Y S @vys]

@aelluswamy I strongly suggest replacing the old autopilot with FSD lite and making it free for everyone. Plus, offering a per-mile pricing for full FSD (say 5c/mile) would make a whole lot more people use it all the time.

[Vinay Y S @vys]

@sriramk that's a good way to source organic text for training ai further! 😂

[Vinay Y S @vys]

shower thoughts: what if ram was disagged and pooled – ram as a service – so it could be plugged into the active device in use. I got my work laptop, home laptop, home desktop, home server, multiple old phones etc lying around. in total I have 100s of GBs of ram around me.

[Vinay Y S @vys]

current ram price trajectory is really killing the vibes. something has to happen soon.

[Vinay Y S @vys]

@elonmusk Please fix parking in an assigned spot in a parking garage, and learning and remembering the right path to the assigned spot.

[Vinay Y S @vys]

e2e latency is sum of ground fiber hop and satellite hop latencies. Satellite closer by 30-40% only cuts that hop. In this case, 1-2ms reduction in a 25ms e2e latency. 5ms is not e2e. They should publish a technical paper.

[Elon Musk @elonmusk]

Starlink V3 satellites have >10X bandwidth of V2 and there’ll be >10X launched, which means >100X more bandwidth.

Also, altitude will be 350km vs 550km, so min latency can be cut in half.

Light travels 300km/ms in space, so physics round trip min latency drops to <5ms.

[Vinay Y S @vys]

@sama solve corruption, pollution, nutrition, health.

[Vinay Y S @vys]

@peyushbansal It should be easy to find out when and why such rules were put up for storefront employees. Do you enforce similar rules for your back office or corporate employees as well?

[Vinay Y S @vys]

I guess the big difference between then and now is how open and accessible the knowledge is and how scalable it is to teach and apply that knowledge+skill to billions of people everywhere consistently. Without that modern engineering would also be indistinguishable from magic.

[Devon Eriksen @Devon_Eriksen_]

The distinction between engineering and magic is entirely a modern linguistic conceit.

To the ancients, a wizard, a magus, a sorceror, was not a man who commanded forces outside the laws of nature. He was a man who commanded the forces of nature, by manipulating them through his understanding of natural law.

But the modern word for a man who commands the universe by understanding its laws is "engineer".

Yes, the ancient sorceror would try to commune with the spirits of the dead, or read the destiny of kings in the stars, or perform fertility rites to make the crops grow, but this wasn't some special supernatural discipline to him.

This was simply his model of how the natural world worked.

He would not have made a distinction between understanding heat and phase changes, and thereby distilling alcohol, and cutting out the intestines of a bird to predict the fortunes of a business venture.

Both, to him, were philosophy and natural law.

But as our understanding of the laws of physics grew more sophisticated, we gradually exiled the term "magic" to that which had not been proven to work, and to that which had been proven not to work.

Were we given the opportunity to take an ancient Egyptian king on a tour of modern society, riding in an electric car, he would remark that we are a rich people, because we have many powerful magicians.

Some of us might hasten to correct him, telling him that there is no magic used here.

But he would not, in fact, be wrong.

[Vinay Y S @vys]

@elonmusk चित्रं देवानामुदगादनीकं चक्षुर्मित्रस्य वरुणस्याग्नेः ।
आप्राद्यावापृथिवी अन्तरिक्षं सूर्य आत्मा जगतस्तस्थुषश्च ॥

[andrew chen @andrewchen]

the magic of cowork and openclaw and other AI products is that they replace our giant row of infinite browser tabs

And lol - no, don't feel guilty, I have too many tabs too. AI makes it so that every workflow that required 4 browser tabs and a spreadsheet is getting collapsed into one AI-native experience

Just as one quick example-
think about how you used to research a person or a company: LinkedIn tab, X tab, Google tab, notes doc, slack open. now one prompt does it in 10 seconds. the "tab count" of a workflow is basically a proxy for how much AI can compress it

if your product eliminates 6 tabs and a copy-paste loop, users will like it. If you can create a whole series of these workflows then your users will absolutely love it. Thus the biggest opportunities are workflows where people currently alt-tab 20+ times per task. Sales, recruiting, research, compliance, procurement. Boring? yes. Massive? also yes. But this is why these agentic tools are going to crush

AI doesn't need to be superintelligent to be wildly useful. it just needs to be good enough to close the tabs

[Anish Moonka @anishmoonka]

Every time you get a cancer biopsy, the lab makes a tissue slide that costs about $5. It shows the shape of your cells under a microscope, and every cancer patient already has one on file.

There’s a much fancier version of that test called multiplex immunofluorescence (basically a protein-level map showing which immune cells are near your tumor and what they’re doing). It costs thousands of dollars per sample, takes specialized equipment most hospitals don’t have, and barely scales. But it’s the kind of data oncologists need to figure out whether immunotherapy will actually work for you. Right now, only about 20 to 40% of cancer patients respond to immunotherapy, and one of the biggest reasons is that doctors can’t easily tell whether a tumor is “hot” (immune cells actively fighting it) or “cold” (immune system ignoring it).

Microsoft, Providence Health, and the University of Washington trained an AI to analyze the $5 slide and predict what the expensive test would show across 21 different protein markers. They called it GigaTIME, trained it on 40 million cells in which both the cheap slide and the expensive test coexisted, and then turned it loose on 14,256 real cancer patients across 51 hospitals in 7 US states.

The results landed in Cell, one of the most selective journals in biology. The model generated about 300,000 virtual protein maps covering 24 cancer types and 306 subtypes. It found 1,234 real, verified connections between immune cell behavior, genetic mutations, tumor staging, and patient survival that were previously invisible at this scale. When they tested it against a completely separate database of 10,200 cancer patients, the results matched up almost perfectly (0.88 out of 1.0 agreement).

Nature Methods named spatial proteomics (mapping where specific proteins sit inside your tissue) its Method of the Year in 2024, and specifically cited GigaTIME in a March 2026 update as a model that “democratizes” this kind of analysis. The full model is open-source on Hugging Face. Any cancer research lab with archived biopsy slides, and most of them have thousands, can now run virtual immune profiling without buying a single piece of new equipment.

[Satya Nadella @satyanadella]

We’ve trained a multimodal AI model to turn routine pathology slides into spatial proteomics, with the potential to reduce time and cost while expanding access to cancer care.

[Vinay Y S @vys]

Software gets denser and wilder: Multi-agent systems layer reasoning, memory, interoperability. Complexity surges—agent sprawl requires tight governance.

[Vinay Y S @vys]

Predictions
Agentic AI boom = massive spike in GPU AND non-GPU compute/storage. Heavy inference + orchestration, memory, observability, and tool execution eat resources across the stack.

[Vinay Y S @vys]

Custom agents proliferate: Individuals and teams build niche, esoteric solutions at scale. Each one demands its own compute, memory, and storage footprint. Distributed digital explosion.

[Vinay Y S @vys]

@elonmusk Money buys time autonomy (capacity). Happiness throughput is still capped by purpose, relationships, health, identity (the real critical path).

[Vinay Y S @vys]

My terminal looks like those in Hollywood movies when multiple sub-agents within Claude code are doing tasks and flooding it with updates that scroll past quickly 😂

[Vinay Y S @vys]

Why does Apple TV voice button not have a fully functioning AI agent? It is such a waste of opportunity.

[Vinay Y S @vys]

@BillAckman All reward programs are based on perception packaging and breakage.
Chasing credit card points rewards is waste of time.
Banks make money and you lose time. For merchants, it is customer acquisition cost, for high value purchases they give you discount if you pay via debit.