I played a round of Geoguessr against it and while it did a shockingly good job compared to what I was expecting, it still lags behind even novice human players.
The locations and its guesses were:
Bliss, Idaho - Burns, Oregon (273 miles away)
Quilleco, Biobio, Chile - Eugene, Oregon (6,411 miles away)
Dettighofen, Switzerland - Mühldorf, Germany (228 miles away)
Pretoria, South Africa - Johannesburg, South Africa (36 miles away)
Rockhampton, Australia - Gold Coast, Australia (437 miles away)
Fentanyl is made from common chemicals that are used in normal industrial processes. We use them for everything from making insulation to medicine. And it only takes a small amount of these chemicals to make a large batch of fentanyl. All the fentanyl produced in a year only takes 1,800 gallons (around 33 oil drums) of chemicals to make.
Noted that production of those precursors has shifted to India.
The fentanyl itself is made in labs in Mexico and then smuggled across the border. It requires no sophisticated lab equipment to make. You can easily obtain everything needed at consumer retail stores and make a batch in a garage. One liter of finished fentanyl is enough to create 50,000 to 100,000 doses.
So if you squeeze the balloon, it just pops up somewhere else. Put pressure on China and India starts supplying the chemicals. Start shutting down Mexican labs and they'll make the stuff in Oklahoma.
Not that these are bad things to do but unless you address the actual demand for the stuff it's going to be nearly impossible to eliminate it.
I don't know anything about Fentanyl, but something about this description seems off to me. If it's as easy to make as you describe, why wouldn't people just make it here already? Why are they taking the risk of smuggling it across borders?
Most criminals aren't capable chemists. Given a laboratory and barrels of precursor chemicals, they couldn't make perfectly legal drugs like ibuprofen either. Most capable chemists aren't willing to risk property seizure and incarceration to make drugs for the black market. The few capable criminal chemists and their laboratories are concentrated in places where law enforcement is weak, like gang controlled regions of poorer countries.
I'm endlessly fascinated by the stories that come out about the execs dealings during that period. How they were offered and passed on being the American distributor for the NES (all because the exec saw Donkey Kong running on a Coleco Adam at CES). [0] And how they funded but f'ed up the deal with the Amiga chipset (it should have been theirs but Commodore stole it at the 11'th hour). [1] Or how they were illegally bypassing DRAM price controls in the Tramiel era (the illegal $$$ is the main reason Atari stayed afloat after the disastrous Federated Group purchase). [2] The list goes on and on.
I'm under the impression that there's a lot of real dirty stuff that's been swept under the rug, maybe now lost to time, as many of the execs are no longer with us. A lot is documented in the book "Atari: Business if Fun". [3] A shame that the follow up book "Atari: Business is War" will likely never be finished. [4]
Mainly because they were the only game in town back then. At the time they were the fastest growing company in history... So many $$$, for a time everything they touched turned to gold. Being first in a new industry, they made all the mistakes that subsequent companies learned from and avoided. For example, putting a textiles executive in charge, treating developers like assembly line workers, etc.
This all laid the seeds for their subsequent implosion... Epic rise, epic fall. I wish someone would make a movie about that story.
First you're successful because you didn't listen to anyone telling you that you were going to fail. Then you fail because you didn't listen to anyone telling you that you were going to fail.
There's a story in business that the CEO that built the company is not the CEO that can keep it running. That definitely seems to be true.
> Every story I hear about Atari is wild. Hard to believe they managed to have the success they did.
It was a different era.
I worked in a mall arcade in the early 90s, and because we purchased arcade games, I had access to the trade shows and various promotional events. For instance, E3 invited me to come out for their first event.
The size of the teams in the early 90s was TINY; I met the dudes who made Mortal Kombat at the AMOA convention, and the entire team was less than ten people. The main programmer had so little experience, he was largely known for doing the voice of "Rudy" in the pinball game "Funhouse."
Basically, the tech community was tiny and the gaming community was a tiny subdomain of the tech community.
Atari's big innovation may have simply been that it was founded in the right location (Silicon Valley.) If it wasn't for that, Steve Jobs wouldn't have worked at Atari. (And Wozniak wouldn't have moonlighted at Atari.)
Atari was the only major gaming company based out of Silicon Valley
A lot of the games of the time were basically just Japanese games that were licensed by US distributors. Pac Man came from Namco in Japan and was distributed in the US by Chicago's Midway, Space Invaders was made by Taito in Japan and licensed in the US. (Also by Midway, IIRC.) "Defender" was one of the first 'homegrown' games in the US that wasn't coming out of Atari in Silicon Valley. (Defender was made by Eugene Jarvis in Chicago for Williams, who later merged with Midway.)
Although Nintendo was NOT based in Silicon Valley, they had the dumb luck of locating just up the hill from Microsoft. If you've seen "King of Kong," the dude from the documentary basically lives halfway between Microsoft in Redmond and Nintendo in Snoqualmie: https://nintendo.fandom.com/wiki/Nintendo_North_Bend
Sega WAS based in Silicon Valley, but their slow decline was arguably due to a political tug-of-war between Sega of America (based in Silicon Valley) and Sega (based in Japan.)
I'm part of a bunch of book collector groups because it's pretty much the only place on the internet to connect with other people around the world about it.
I run every query I do through all the major models, up to 10 of them at this point.
Benchmarks aside Gemini 2.5 Pro is a great model and now often produces the best code for me but it's not notably better than any of the other frontier models in my testing each of which tends to have their own strengths and weaknesses.
And Google's wrapper around Gemini is easily the most frustrating of any of the major AI companies. It's content guardrails are annoying and I just learned yesterday it won't let you upload json files for whatever reason (change the extension to txt without modifying the contents in any way and it works just fine).
Gemini 2.5 Pro does this annoying thing where it decides to refactor every part of your code even if you didn't ask, and also it outputs way too many damn comments on almost every line in the style of:
// Increment variable by 1
I find Claude 3.7 better at following instructions, even though the solutions it comes up with may not be the best at times
The locations and its guesses were:
Bliss, Idaho - Burns, Oregon (273 miles away)
Quilleco, Biobio, Chile - Eugene, Oregon (6,411 miles away)
Dettighofen, Switzerland - Mühldorf, Germany (228 miles away)
Pretoria, South Africa - Johannesburg, South Africa (36 miles away)
Rockhampton, Australia - Gold Coast, Australia (437 miles away)
reply