eitally

At least in the case of Google, the pace of their buildout is very nearly being matched by the growth of Cloud consumption (and related contracts with companies like Anthropic & OAI). Cloud had >100% y/y growth this quarter to >$30b rev. That's insane, and it's quite profitable. But consumption revenue is always a lagging indicator of capital investment, and it just so happens that this business cycle is demanding a scale of capex heretofore unimaginable. Institutional investors seem to have been happier with business as usual (big annual stock buy-backs, predictable cloud/ads/youtube growth in the high-20% range, EPS in the $5-9/share/qtr range, and an enormous pile of cash). This is the first time since 2004 where buying GOOG really feels like placing a bet rather than buying a blue-chip. To that point, this is also why GOOG has been treated like an AI company for the past 18mo or so, rather than a media & infrastructure company, and what justified the dramatic increase in it's P/E.

Taking a longer view, I have confidence things will work out well and this dip is a great buying opportunity, but I can also understand why the market several reasons to be skittish.

Even in California it's legal to install non-grid-tied rooftop solar. It requires permitting but does not require an interconnect agreement from the grid electrical utility. We're "remodeling" our house in San Jose currently and the resulting property (includes new ADU) will have 1) old panels (NEM2) grid tied for main house, 2) new panels battery tied for main house & EV charging, 3) new panels (NEM3) grid tied for AGU. PG&E does not care about the second item, provided they are installed per code.

No. It means customers can turn booked revenues into billed revenue by consuming OAI models on GCP (pass-through contracting). OAI will get their money, too, but Google will charge the customer for actual platform consumption. This was previously missing and GCP customers who wanted ChatGPT or Codex were limited to just making API calls from their GCP projects .

It's showing up for enterprises (and ISVs) who need capacity from cloud vendors and can't get it. Sometimes there are some GPU shapes available, but not always what you either need or want to pay for (e.g. a TPU vs a V100 vs a B300), and also not always in spot instances (frequently not).

The hyperscalers are all so oversubscribed that they require commits from enterprises so if you only have a transient need you might be out of luck [from your vendor of choice]. Additionally, not all GPU service providers are equal. Semi Analysis has some great articles explaining why.

It may not be the most prescient or innovative sci fi series, but I read The Three Body Problem a few years ago and it's really stuck with me. Specifically the description of how humanity reached an inflection point in math & physics research that almost immediately unlocked novel innovations in applied science ... which then made space exploration & settlement practical. Moreover, that the antagonist society in the series decided their absolute best bet at avoiding complete annihilation was to hamstring humankind's scientific research -- and that's it.

It feels like frontier models are starting to approach an asymptote where what you suggest could become reality, and we'll begin seeing not just AI research but AI-designed & led applications of that research into fundamental parts of our society & culture. For better and for worse.

I've had pretty good luck using Google Appscript to either access web APIs or screenscrape, then parse into a Google Sheet (which I either use as-is or as the source to populate a web app).

Totally off-base projection based on your very small sample size.

I've spent quite a lot of time with corporate counsels over the past dozen years and one of my consistent findings is that they -- except in areas of liability and intellectual property -- almost always defer to the sponsoring business executive in defining terms and specifying the details of the contract. They are present primarily to keep biz leaders and their counterparties operating within an acceptable risk profile but usually don't have strong opinions on much of anything else.

So, to me it isn't curious that your experience with Claude has been positive. Provided you know what you're doing and you aren't asking Claude to go too far outside the expected norm, there wouldn't typically be a compelling reason for most contract attorneys to push back.

I think the audience for home automation is U-shaped. There are folks who are neck deep in it and care deeply, and others who couldn't care less, but not many in the middle. The middle 90% is decently served by Apple, Google and Amazon, too, so there's no need for any real passion or exploration by most anyway.

We're building a custom home and it's only been through this process that I've taken an interest in smart home / home automation. Why? Mostly because

1) building speakers into your walls & ceilings means you need to control their input somehow, and your gear is going to be in a wiring closet

2) if you don't want Alexas or Nest devices sitting around on horizontal surfaces, your voice control options are limited

3) you don't want your security camera footage stored in the cloud

4) you do want a single pane of glass for visibility into things like your HVAC, your irrigation system, your security system, your garage door status, and your AV gear.

Home Assistant is tremendously helpful as a FOSS alternative to Control4, Crestron, Savant, Lutron, and Josh.ai, but you can't entirely escape the "big tech" part of the ecosystem because each of the primary systems (say, your AV streaming or your power shades) will still require use of the OEM's product. Many of which have integrations with or APIs exposed to Home Assistant, but not all.

There are some quality of life improvements available through automation, but most of it is incremental and only enthusiasts would really care.

San Jose. My wife and I have lived in the same house for 10 years and know about 80% of the neighbors in a two block radius. We also probably know 50 additional families within a square mile as a result of our kids going to public schools here. This is essentially just like how it was when I grew up in the 80s, in a rural East Coast City of about 75k.

This, but worse, happened with my inlaws when they purchased a new Lexus (IIRC, a 2025 RX350). It has whatever kind of built in navigation & traffic feature but they wanted to use Android Auto. You can do that, but not until after you go through Lexus' vehicle/driver registration process to enable their built in system. If you don't, not only will neither AA nor CarPlay work, but if you try to use the built in nav it will fail with a superimposed modal after a couple of minutes to nag you into creating an account.

That isn't what the OP was alleging. Of course they store data -- encrypted both in transit and at rest, and you can even activate client side encryption (https://knowledge.workspace.google.com/admin/security/about-...) -- but that doesn't mean they exfiltrate, use or share your data for their own purposes.

They don't, they never have, and they won't (unless there is a blatant, illegal policy violation and lapse of controls internally). This is very similar to the controls and policies in place governing FedRAMP contracts or the control of PCI/financial or HIPAA/healthcare data.

None of those have anything to do with Google for Education contracts, and I hope you realize that.

Education contracts do not allow Google to use customer data for advertising.

When we moved to the bay area in 2015 (south bay), I was absolutely stunned at the variety of residential architecture, having come from the mid-Atlantic where everything tended to be uniform based roughly on construction date/period. Since land is exorbitantly costly, homeowners tend to build what they like, no matter whether it fits the neighborhood aesthetic or not.

I recommend Doctors of Running (website & podcast). They're a group of running PTs and each of them has different needs & preference. One in particular likes less cushioned and more structured shoes. There are ample options out there.

That's just a bad salesperson.

Conventional wisdom among running coaches and orthopedists these days is that the most important thing is to run comfortably, and that just because something looks "weird" doesn't necessarily mean it's bad or wrong. Yes, it may be inefficient, but it probably isn't hurting the runner, either.

Shoes salespeople wield huge influence over buying decisions, especially if they come off as authoritative.

My son works in a running shoe store that also does gait analysis and they default every new runner into something stable-neutral like a Brooks Ghost, Saucony Ride, or NB 880. They never default with Nike or Adidas because, although both brands make some excellent running shoe models, neither have anything that's in the same category (if Nike had a shoe that was a mix between the Structure & Pegasus, that would be id. If Adidas had something like the SL2 but with a little more structure, that would, too.). The problem with both Nike & Adidas is that they cater to a million different niche audiences, and while both have some amazing running shoe options at the top end (e.g. Vomero Plus for cushioned daily, Alphafly / Streakfly / Vaporfly for racing, etc) their shoes targeting casual runners doing 2-5mi slow runs a couple days a week are lacking. Yes, you could go with a Pegasus, and many do, but although that's a general purpose stable-neutral running shoe, there are way better options from other brands (including Hoka & On, but also Brooks, NB, Saucony, Asics, Topo, Altra, Mizuno, and more).

Nike are too narrow for me. I alternate between Inov-8 and Topo for trail shoes. I was really happy with an older pair of Adidas Terrex Pro, but haven't tried the new Agravic options in several years.

Trail running is highly dependent on the type of trails around you. I'm in the bay area and almost everything here is dry, dusty and hard. You can generally get by with structured road running shoes if you want to, and don't need much support beyond that. I would not recommend the same thing to east coast trail runners, or most northern Europeans.

I think if you looked closer at runners you'd see many wearing Adidas, Asics, Brooks & Saucony, too. NB may be a mom/dad fashion brand with models like the 99x, 540, etc, but they're a default starting point for many runners because they're one of very few options that have wide an extra-wide across almost their entire lineup of running shoes.

You mention a handful of specific shoe models but none of them are running shoes (not even the "Night Jogger").

For my daily shoes, I have been wearing cushioned running shoes for the past ~10 years. For my daily trainers, I currently wear On Cloud Eclipse. For workout shoes I alternate between NB SC Trainer and On Cloudmonster Hyper. For trail running, Topo Vista or Inov-8 G280 (now called Trailyfly). For racing, NB SC Elite.

I also have a pair of Sambas for fashion; they're too narrow for anything else. I do light hiking in approach shoes (La Sportiva TX4 -- I ADORE these, to the point I bought their waterproof mid-height version for more extreme hiking), too!

(I am a runner, my son is a high school runner and works in a running shoe store. I'm deeper into the running shoe market than most.)

This is why it's not recommended to use providers below the Silver level in Semianalysis' rating index.

https://newsletter.semianalysis.com/p/clustermax-20-the-indu...

As an engineering leader back in the day (2000-2015) I considered it terrific value to license Telerik and others. In fact, it looks like Progress has acquired a large catalog of the most common .NET dev tools over the past ten years.

They offer inference, too. I don't know anyone who uses this service, but presumably it must be reasonably popular.

This has become a common pattern in my Claude & Gemini usage. Always require citations, then check those citations to validate they actually contain the information/data the LLM's output claims. Claude, in particular, seems to make massive logic jumps and trust tertiary data sources way more than it should.

Honestly, there's just a huge amount of thrash across the cloud & SaaS industries as a whole. It's not a lot better at Google (way behind OAI & Anthropic on coding & local AI harnesses), Microsoft (no meaningful frontier model R&D to speak of, and a constantly eroding Windows business), the neoclouds (impossible to source adequate infrastructure and are existentially threatened by the hyperscalers (especially Google) if they can't keep up with demand, chipmakers (NVIDIA vs AMD vs Qualcomm vs Apple vs the long tail of specialty shops like Cerebras), data center pure plays (legislative & community pushback, natural resources, infrastructure availability), and the SaaS folks ("let's see if AI can do it").

It's both a great time to be gold mining in tech, and also a terrible time to be a bit employee.

Maybe, but GPU is just one aspect of NVIDIA's dominance. If you are buying Vera Rubin GPUs, you're getting an NVL72 rack, which is only one of several racks that you're probably buying. You'll also need your NVIDIA racks with NVIDIA networking & storage gear, too. At the end of the day, they're "vertically integrated" for your accelerated computing data center (e.g. the "AI Factory"). This doesn't even count the software layer, where CUDA + CUDA-X (not to mention the software for all the sysadmin pieces) has a huge first mover advantage over anyone else.

How did you effectively download 500GB of photos? Did you just use Takeout?

I think it's arguable that this "intuition" you call knowledge could very easily also be called "opinion".

The interesting thing is that China began as an agrarian economy and quickly modernized via influence/direction of the CCP.

The United States used to have a dominant middle class that was geographically distributed (cities & rural areas inclusive), but the advent of the tech economy has also been having a similar effect here as it did in China: massive wealth accumulation in Tier 1-3 cities and everyone else being largely left behind.

It'll be interesting to see if there's convergence in the next decade or so, especially with the GOP reducing regulation and increasing the explicit capitalist priorities here. TBH, though, as a generally well-informed political outsider, the feeling I have is that the US government is slow, bulky, inefficient, balkanized and overall poorly run compared to the Chinese government. We'll likely either slowly improve authoritarian efficiency and become more like the CCP ... or we'll pivot left and move more toward the EU model, but we're floundering around right now paying lip service to both.