

oh no there are other answers other than this… how silly :(
oh no there are other answers other than this… how silly :(
Assuming you shot this on a phone, it’s NUTS how good mobile cameras / post-processing has gotten in the past couple years. Fantastic pic.
Probably. I write half my comments drunk, so I wouldn’t use them as a basis for ESL learning 🙃
It’s a good catch!! Apologies for any confusion.
I agree wholeheartedly. Whenever some is doing app design, I always tell them - just because the PNG submission recommendation is 1024x1024px, there’s only one time people will ever see it at that res (in the store); after that, it gets scaled down as small as 120x120 for Home Screen use (or 58x58 in settings).
There’s a reason boring, flat vector design wins - fast, universal legibility.
It’s nice art… for anything mass adoption, I always recommend flatter/less detail. No one like FB Buck Alegria, but man does it work for global tech deployments.
This is a really good oversight (see: insight, overview, etc). Honestly, for anyone actually interested in this stuff and what makes the internet tracking/advertising machine tick, take some of the HubSpot Academy’s courses. There’s definitely other courses out there, but the HubSpot ones are all free, and the topics aren’t hard once you get immersed in it.
Plus afterwards you can put the faux-certs on your resume and knife fight with the 20,000,000 other adtech people that just got laid off.
These look perfect. Charred corn? Pickled onion? Cotija? Sign me up.
Come to Canada. Eat the curds. Alter your mind.
This is actually a proven idea in net new real estate development involving wetlands and protected acreage; you can build on wetlands, but for every acre you displace, you have to create two acres, and both the plan and results are audited.
To your point, the end result of this - in many cases - is to simply build elsewhere due to the considerably higher costs. I think a model similar in energy would pay dividends rather quickly - most likely, we’d see Shell, EM, CP, etc. rapidly transition to renewables from an imposed cost perspective.
You bring up lobbying - definitely the major hurdle. Fortunately, if you go read these guys 10k’s, I think the shift is inevitable, they’re just artificially pumping the brakes to adhere to some kind of amortisation timeline of investments they’ve already made… which unfortunately, is super frustrating.
I mean, he sucks, but I also don’t disagree that having a bridge solution during OPEC fuckaround times is a good idea. Supply cuts aren’t fun, period, especially when we can’t influence them whatsoever (since it seems dollar hegemony is going bye-bye at some point):
There’s probably a decent solution to temporarily boost and maintain domestic oil production while also feeding green R&D, whether that’s in the form of a tax, credit, government package, or full-blown turn-Shell-into-an-SOE; all of its years vs. months, I think. Bottom line, I think it’s a more complex solution than “fuck this guy”.
Amazing.
I went to Saint Malo as a child, and still have vivid, core memories of it… one of the most beautiful places I’ve ever been. Good way to unlock a love of old, walled, coastal cities.
Others have basically captured it, but my read is a massive change in the overall risk profile held by venture capital firms. The time of reckoning has come, and it’s time for everyone’s (or at least VCs’) favourite three letters: ARR (Annual Recurring Revenue).
The last twenty years, we’ve seen this sort of spray-and-pray model, where 99 bad investments could be offset by 1 “unicorn”. The risk appetite seems to have shifted largely because 1.) there’s a higher volume of early stage concepts (so there’s more bad ideas), and 2.) there’s either fewer unicorns, or the unicorns that mature are ultimately less valuable.
Crunchbase put out a good analysis of the current trend of global venture dollar flow:
The Party’s Still Over: The VC Downturn In 6 Charts
You can read news from various outlets - some say it’s a post-pandemic correction. Some say it’s because labour is too expensive. But the bottom line is that VCs aren’t willing to spend money on “users-in-lieu-of-revenue” like they once were, and I honestly don’t blame them. There were a lot of really, egregiously stupid ideas coming out of SV, and their wax wings melted. sad_trombone.mp4
Adam Kotsko summed this entire phenomena up nicely:
I don’t agree with this statement. I think “intelligence”, however you define that, has fit pretty cleanly to a Gaussian distribution since the dawn of man. If anything, I think advances in nutrition, preventative healthcare, and access to information has driven pretty significant negative skew. I don’t have anything to back up this claim, but my guess would be that median intelligence has actually increased - it just may not seem as such since the population continues to rise, so the raw number of dummies seems overwhelming.
But hey, who I am to define what’s smart? Maybe an inflatable hot tub, 30 rack of Busch, RAM 1500 on an 84 month note, zero turn mower, DirecTV with Fox News and the funds to pay for it all is the real secret to a happy life. I’m just someone blabbing about nothing with a bunch of Reddit exiles.