Anirudh Vemprala retweeted
The late Karl Weick captured this beautifully: “Argue as if you are right. Listen as if you are wrong.”
When someone disagrees with us, our first instinct is often to prove that we are right and they are wrong.
Unfortunately, this instinct can get in the way of learning.
One of the most useful habits I have tried to develop is making positive assumptions about other people. Unless proven otherwise, I try to assume that they mean well. I try to assume they are intelligent. I try to assume that their perspective comes from experiences that may be different from my own. Sometimes I fail, if I remember I try.
People do not have to be wrong simply because they are different.
When we stop focusing on winning the argument, we create an opportunity to understand. We begin to see the world through another person's eyes. Sometimes we learn something new. Sometimes we discover common ground. Occasionally, we even change our minds.
Listening does not mean agreeing. Respect does not require surrendering our beliefs.
It simply means being open to the possibility that another person may see something we do not.
And who knows? We may end up agreeing with them after all.
Life is good.
Marshall
Anirudh Vemprala retweeted
Totally insane levels of corruption and abuse of power are swirling all around us and my sense is that as angry as voters are about the cost of living they mostly are so awash in cynicism about politics that they don’t understand how egregious Trump is.
Billions of dollars’ worth of contracts predicting oil would fall and the S&P 500 would rise were placed about 15 minutes before Trump announced a U-turn on Iran in a post on Truth Social. CFTC staff began to investigate-- then were told to lay off. bloomberg.com/news/features/…
Anirudh Vemprala retweeted
This interview could be used for ‘Agentic Commerce is a Mirage’ bingo.
✅ Asserts that agents can optimally monetize with transaction fees (fees are sub-optimal relative to ads at scale.)
✅ Highlights restaurant reservations as a core agentic use case (in-demand restaurants *benefit* from controlling scarce inventory / availability. This is also likely a tiny market, and consumers may not be willing to pay for this functionality.)
✅ Presumes that all retail platforms must give way to monolithic agents that handle a wide range of tasks or face extinction (eCommerce is concentrated, retail platforms offer their own agents, and the independent agent space is crowded. Agents need retail cooperation more than vice versa.)
Instinct CEO Noah Shinn says agents will let restaurants give tables to birthdays and anniversaries over whoever clicks first
"And traditionally, because of the way that the internet has worked, at least for reservations, it's been first come first serve. So anyone that comes in, no matter how important or how not important it might be, whoever gets in first gets the reservation."
"But what if you had the ability for the agent to be able to communicate on both sides, communicate, hey, this is important. This is actually the person's spouse's birthday, the 30th birthday, a very special event is coming up."
"And the restaurant too can say, let's actually prioritize birthdays or major decade birthdays or major anniversaries or these certain people, I don't know, whatever it might be, right?"
"So now with the ability, with all of the logistical burden of having to describe exactly what your case is and what's happening and why it's important. Now we have the ability to perfectly match what does the restaurant want? What does the user want?"
"And that will result in more special events for the users being able to actually have a spot in the restaurant."
Anirudh Vemprala retweeted
This one got to me
Perhaps because he’s not an anon, but the favorite to win statewide office in Texas.
Perhaps it’s because the kids in the picture could have been me 25 years ago.
I just donated to his opponent and hope you’ll join me
secure.actblue.com/donate/ro…
Anirudh Vemprala retweeted
100% agree. Using LLMs to write for you is going to shrink your capacity to verbalize your thoughts and generate sharp responses when challenged. Much better to use them to stress-test your views and reveal your blindspots.
Anirudh Vemprala retweeted
Most video AI processes a file.
Jockey processes a corpus.
That's the whole difference.
Anirudh Vemprala retweeted
This is amazing. The Atlantic posts JD Vance’s oped from 2016.
“A few Saturdays ago, my wife and I spent the morning volunteering at a community garden in our San Francisco neighborhood. After a few hours of casual labor, we and the other volunteers dispersed to our respective destinations: tasty brunches, day trips to wine country, art-gallery tours. It was a perfectly normal day, by San Francisco standards.”
Anirudh Vemprala retweeted
The most important lesson from the Cabo Verde match last night was to dare to be great!
Believe in yourself to be the very best, put that work in, and ignore what anybody else says.
The most important thing to notice about this team isn’t the epic World Cup run that nobody expected, or the insane performance of 40 year old keeper Vozinha, or the crazy strike to tie for a second time last night from Lopez Cabral.
The most important thing is the sadness, frustration and grief expressed by the players after the game.
Why?
Because these guys are winners and have a winning mindset.
Nobody gave them a shot against Argentina.
But their attitude wasn’t “oh we tied some good teams and we were lucky to be on the field with Argentina and give them a run.”
Not at all.
Their attitude was “I thought we had it, I’m so frustrated and disappointed right now.”
These boys came to win. They knew they could win. They had no sense of satisfaction that they just made it to this point and gave Argentina a run for their money. They played to win.
That is what it means to dare to be great. You have to believe you can win on the biggest stage imaginable.
That’s the lesson.
So, we are all proud and we celebrate this epic run, but please also notice how the players felt after the game because they weren’t celebrating - they were disappointed because they believed in themselves to win regardless of what anybody else thought.
That’s the dare to be great mentality!
Yes! We have a business and team there for @alphacityinc and I can assure you this will have a permanent impact that makes the country dare to be great
A big part of the battle for developing Africa is just getting the confidence up to dare to be great, and this World Cup run will impact not just CV but Africa overall
I’m here at Afronation right now in Portugal and the energy was electric watching the game last night, even drawing people to the football screen over Burna Boy’s headline show on the main stage
Everyone is so proud of them 🇨🇻
That strike to tie the second time was just insane
Anirudh Vemprala retweeted
JUNE 1, 2001
🚨MICROSOFT CEO: OPEN-SOURCE OPERATING SYSTEMS ARE DANGEROUS
$MSFT CEO Bill Gates told lawmakers that open-source operating systems such as Linux are "going down a very dangerous path." Transcript below ⬇️
"""
The scaling of open source operating systems, I think it's going down a very dangerous path. And again, if the path continues, I think we could get to a very dangerous place.
I think it's worth saying some things on Linux that are clear to all the experts, but I want to make sure is understood by this committee, which is when you control the operating system and you're shipping it, you have the ability to monitor its usage. It might be misused at one point, but then you can push an update. You can revoke a user's license. You can change what the system is willing to run.
When an operating system is released in an uncontrolled manner, by some guy compiling his own kernel in his basement, there's no ability to do that. It's entirely out of your hands. And so I think that should be attended to carefully.
There may be ways to release software open source so that it's harder to circumvent the licensing, but that's a much harder problem, and we should confront the advocates of this with that problem and challenge them to solve it.
Finally, I'd say open source is a little bit of a misnomer here, right? Open source normally refers to smaller developers who are iterating quickly, and I think that's a good thing. But here we're talking about something a little bit different, which is a more uncontrolled release of larger systems by, again, to your point, Senator Hawley, like much larger entities that pay tens or even hundreds of millions of dollars to develop them. I think we should think of that in a little bit of a different category, and their obligations in a little bit of a different category.
"""
🚨ANTHROPIC CEO: OPEN SOURCE AI IS GETTING DANGEROUS
Anthropic CEO Dario Amodei told lawmakers that open-source AI is moving down a “very dangerous path.”
His warns that once powerful models are released openly, companies lose the ability to monitor misuse, revoke access, or update safety guardrails.
Anirudh Vemprala retweeted
A doctor on supplements ...
My conversation with Dr. Giulia Enders, the leading authority on the gut microbiome.
I was curious about how to build, maintain, and restore gut health, the link between the gut and the brain, the truth about prebiotics and probiotics, and the report card your bowel movements give you every day.
This episode is fast-paced and packed with practical advice, including one simple thing that changed how I make rice forever.
Enjoy!
(Includes paid partnerships)
Anirudh Vemprala retweeted
How to keep AI spend flat while token usage grows exponentially: Not with friction and spend alerts. With better defaults, routing, and caching.
Better Defaults (not Usage Caps) – Engineers can choose any model they want, but defaults matter. We’re experimenting with defaulting to open weight models like GLM 5.2 and Kimi 2.7 through our LLM gateway, while still encouraging engineers to choose the right model for the task. 91% of our employees were never hitting their usage caps, so instead of lowering caps and driving up alerts, we're moving to cheaper defaults. Note that code reviews use a diversity of models, so they can check each other's work.
Better Routing – In our custom harnesses, we preprocess prompts and route to the best model for the job, considering cache hits and model pricing. For instance, you may want a frontier model for planning, but not for execution where they can be overkill. Ultimately, humans shouldn't be choosing models - AI can automate this task.
Better Caching – Cache misses are the easiest way to drive your cost up. All of our requests are cache aware, so we’re reusing a warm cache wherever possible. For example, our cache hit rate went from 5% → 60% in LibreChat once properly implemented.
Keep Context Lean – Start fresh sessions when switching tasks. Scope file context narrowly. Disconnect unused tools. Don't just compact. The goal isn't fewer tokens used, it's fewer tokens wasted.
Better Visibility – Our engineers can use as many tokens as they want, from whatever model they want, but we’ve made usage visible – and the more you spend on AI, the more impact we expect.
The goal isn't to suppress usage. It's to build the infrastructure that makes exponential growth sustainable.
Putting this into practice has cut our AI spend nearly in half, while our token usage continues to grow.
Anirudh Vemprala retweeted
One of the ways scaling test-time compute can benefit people most: have reasoning models think really hard about rare undiagnosed diseases.
Today we’re sharing published evidence that this can work, in some of the most difficult pediatric cases!
Anirudh Vemprala retweeted
The reflecting pool is now green, the color of all the money wasted on it. Painting the bottom has apparently heated the water and accelerated algae growth. Pure Genius.
people.com/reflecting-pool-g…
