Micro

Perplexity’s Comet is the AI browser Google wants

Comet also comes with an AI assistant built in, similar to the Gemini integration that Google is testing in Chrome. Selecting the Assistantbutton in the top-right corner of the browser will open up a sidebar with a chat interface. From here, you can type in a query or use voice mode to chat about different topics, as well as ask specific questions about the webpage you’re on.

I wrote About AI browsers some time back. I continue feeling the same way about them.

Uber’s Robotaxi Is No Quick Delivery

We have not fully contextualized the impact of the gradual automation of our everyday life and how much it reduces economic activity. Waymo’s driverless profits flow mostly to its investors, employees, and eventually Google’s shareholders. The local economic impact is close to zero, barring a few taxes. 

Humans buy coffee, gas, and stay in the city. They even pay taxes on their income. They support the local ecosystem. A self-driving car company has none of those inefficiencies. Good for profits, not so much for the local ecosystems. Others see Waymo’s success and want the profits, just as fast-food chains want robots flipping burgers.

This is something that I keep wondering myself. If AI/robots replace the human workers, where will the humans get the money to buy the food or service or whatever.

Tasks and projects

Art is a project. Connection, community building, counseling–all of these are projects. When our work is project-focused, we’re not a cog in a vast machine. Instead, we’re a contributor with agency, someone who is working with and for the agenda we’ve agreed to.

The Bad bosses try to have it both ways. They are stingy with agency, authority and compensation, and insatiable when it comes to effort. But smart leaders understand that given the chance, most of us would love the chance to be seen, to contribute and to be part of something.

The Sputnik vs. Deep Seek Moment: The Answers - Marginal REVOLUTION

Zero sum thinking fuels support for trade protection: if other countries gain, we must be losing. It drives opposition to immigration: if immigrants benefit, natives must suffer. And it even helps explain hostility toward universities and the desire to cut science funding. For the zero-sum thinker, there’s no such thing as a public good or even a shared national interest—only “us” versus “them.” In this framework, funding top universities isn’t investing in cancer research; it’s enriching elites at everyone else’s expense. Any claim to broader benefit is seen as a smokescreen for redistributing status, power, and money to “them.”

If there is fixed growth, then, people would think that if someone else is growing that means they are growing at our expense.

Reflections on OpenAI

I have been reading so much news these days about Meta taking AI talent from OpenAI and other companies, it was fun to read this little tid-bit here:

When it comes to personnel (at least in eng), there's a very significant Meta → OpenAI pipeline. In many ways, OpenAI resembles early Meta: a blockbuster consumer app, nascent infra, and a desire to move really quickly. Most of the infra talent I've seen brought over from Meta + Instagram has been quite strong.

There are other fun insights here as well. It’s seldom we get these types of look inside these companies.

How large models are trained (at a high-level). There's a spectrum from "experimentation" to "engineering". Most ideas start out as small-scale experiments. If the results look promising, they then get incorporated into a bigger run. Experimentation is as much about tweaking the core algorithms as it is tweaking the data mix and carefully studying the results. On the large end, doing a big run almost looks like giant distributed systems engineering. There will be weird edge cases and things you didn't expect. It's up to you to debug them.

Read this in full.

My Routine - Hugh Howey

We often don’t even wait to get hungry. We wake up, and it’s breakfast time. We eat a full meal, or some junk like a bowl of cereal. We crush a large coffee full of milk. We snack on something before lunch. We eat lunch, because that’s when we have a break in the day. More snacking in the afternoon. A full dinner. Snack before bed. Any slight hunger pain is a mere itch compared to the real deep hunger we are designed to experience before getting a meal.

The first thing to learn is that hunger should not equal panic. Instead, hunger should be celebrated as a sign of a healthy, functioning body. Sit with the hunger a while. Learn to associate hunger not as something wrong, but something right.

There are many good points in this post. Things that I agree with.

It motivated me to finally write about my workout routine too.

Google’s curated AI ‘notebooks’ talk you through topics from parenting to Shakespeare

The featured notebooks include original text from the source material, whether it’s a book, play, newsletter, or online article. NotebookLM automatically summarizes this information and comes preloaded with notes about the topics discussed in the source material. You can also interact with NotebookLM’s AI chatbot to ask questions about the information, as well as listen to pregenerated Audio Overviews, the podcast-like discussions featuring AI “hosts.”

The NotebookLM product famously came from a Googler’s 20 % time. I have not used this product, but I have found myself spending more time with voice (speech-to-text and audiobooks ).

This seems like a good idea in that aspect.

This new announcement seems like a good idea too, a natural evolution of the product.

There maybe just a little glimpse of the future AI-fied world here too, with individual creators creating things for the chat bots.

I have noticed that I tend not to finish a thought later if I did not write anything about it when the thought came. This burning desire to write, when I actually write something, comes to me once, and if I let it pass, if I get into a state where I am not excited about the thing, then I just am not able to write about it later.

In those scenarios,

1. I need to either scrap the note altogether, or
2. Get excited about the idea again.

There is a third option too, but what's the point of writing a half-assed thing?

Average age of cars in Finland nears 14 years amid sluggish sales

The average Finnish car is 13.6 years old — compared to 11 years in Sweden, 9.6 years in Denmark, and 11.1 years in Norway.

I am a bit surprised by this. I see newer cars on the road. Or maybe what I see is well-maintained cars. From time to time I do see some old cars too though. Super old Yaris and so on. I guess there is a market for those.

I recently bought a car. It’s a 3 year old Qashqai. It’s good to know there’s a market for selling it again after I’m done.

Using AI Right Now: A Quick Guide

For most people who want to use AI seriously, you should pick one of three systems: Claude from Anthropic, Google’s Gemini, and OpenAI’s ChatGPT. With all of the options, you get access to both advanced and fast models, a voice mode, the ability to see images and documents, the ability to execute code, good mobile apps, the ability to create images and video (Claude lacks here, however), and the ability to do Deep Research. Some of these features are free, but you are generally going to need to pay $20/month to get access to the full set of features you need. I will try to give you some reasons to pick one model or another as we go along, but you can’t go wrong with any of them.

1. Pay for subscription
2. Use deep research for trickier stuff
3. Use higher model for complex queries (coding, etc.)

The Last of Us Part II’s new mode puts the story in chronological order

The studio has released a new, free patch for The Last of Us Part II Remastered that lets you play the game in chronological order. The mode takes Abby and Ellie’s stories and interleaves them in chronological order rather than the nonlinear story of the original version of the game, which used timeskips and flashbacks.

This might be fun. I am imagining that there would be cuts from one piece of action to the next.

I just don’t have time to play games anymore.

Gmail’s new tab is made for unsubscribing from emails

The view will show you who’s sending the most emails and exactly how many messages they’ve sent in the past few weeks so you can be better informed about who’s clogging up your inbox the most.

I use gmail’s current implementation of this aggressively. Anything that I do not want another email from, I immediately click on unsubscribe.

I get so many spam type emails that tracking any useful communication has become a challenge. Inbox zero is just not possible.

This would be a welcome addition.

Teaching ChatGPT how to work with me

• You want me to think first, not code.
• You want me to challenge assumptions, including my own.
• You expect me to give up on a theory when the evidence contradicts it, and look for more plausible explanations.
• You're a deeply experienced debugger (almost 50 years), and you know how important it is to notice the obvious.
• You're not here for code generation — you're here for sharp, collaborative thinking.

Finland backs Nokia-led plan for AI gigafactory

The European Commission’s InvestAI initiative is targeting 200 billion euros’ worth of investments in AI and high-performance computing (HPC), including a huge push for AI infrastructure. The petascale supercomputer Lumi is partly funded by the Union’s EuroHPC Joint undertaking.

I was just reading about this yesterday, the different types of funding government can do to guide deployment.

“Now is the right time to influence the development of the European artificial intelligence infrastructure,” she said in a statement last month.

The government bills Finland as "an ideal location for an AI gigafactory, largely due to clean energy grids and land availability".

I wonder how much new electricity is coming up in Finland, so that there is no impact on electricity prices for normal consumers because of all these data centres that are coming up.

Cyberpunk Edgerunners 2 will be even sadder and bloodier

I loved Edgerunners. It was a big reason why I eventually got around to playing Cyberpunk 2077. Cyberpunk 2077 was famously shit at launch, but they fixed it over the many years. The main complaint remained, however, which I felt too, that the final, no way back from here, quest line came too soon.

The quest lines in general in Cyberpunk were the typical CD Project Red quests. There was no good or bad in them. Every choice has consequences. Mostly shitty consequences.

That was what happened in Edgerunners. Something similar (worse) would happen in Edgerunner 2.

Looking forward to this.

Why the AI revolution needs tollbooths

AI web crawlers had begun inundating news and information websites with thousands of requests a day compared to the handful they typically saw from search engines. Not only was the explosion in traffic ballooning hosting costs for these sites, the bots supplied zero traffic to them in return. Web traffic in exchange for permission to crawl has been one of the unseen foundations of the internet economy for a generation.

This sounds like a good idea. The AI companies pay if they want to access your site. It needs to happen at scale, a new way for people to get paid for what they write/make.

Today, Cloudflare Is Blocking AI Crawlers by Default

Last year, internet infrastructure firm Cloudflare launched tools enabling its customers to block AI scrapers. Today the company has taken its fight against permissionless scraping several steps further. It has switched to blocking AI crawlers by default for its customers and is moving forward with a Pay Per Crawl program that lets customers charge AI companies to scrape their websites

Cloudflare has scale. This might be the start of the toll era on the web.