There’s this saying that humans are terrible at predicting exponential growth. I believe we need another saying, those who expect exponential growth have a tough time not expecting it.
It’s not under a decade for ai to get to this stage but multiple decades of work, with algorithms finally able to take advantage of gpu hardware to massively excel.
There’s already feeling that growth has slowed, I’m not seeing the rise in performance at coding tasks that I saw over the past few years. I see no fundamental improvements that would suggest exponential growth or human level of performance.
I'm not sure if there will be exponential growth, but I also don't believe that it's entirely necessary. Some automation-relevant performance metrics, like "task-completion time horizon", appear to increase exponentially - but do they have to?
All you really need is for performance to keep increasing steadily at a good rate.
If the exponential growth tops out, and AI only gains a linear two days per year of "task-completion time horizon" once it does? It'll be able to complete a small scrum sprint autonomously by year 2035. Edging more and more into the "seasoned professional developer" territory with each passing year, little by little.
I find it interesting you are blaming the NYT on this and not ChatGPT for keeping these logs in the first place. If openAI didn't keep logs, then there would be nothing to search, and a more harmful actor couldn't accomplish something far more nefarious. Saying that there could be confidential information in the logs, so that means we shouldn't access it, should also mean the logs shouldn't be kept.
Byd recently came out saying the hyper competitive landscape and low prices needs to end soon. The Chinese government is propping up a lot of their auto industry right now. So some protectionism is needed if you don’t want one of the last bits of manufacturing strength to disappear in the US.
Genuine question, we have manufacturing strength in the US auto industry?
Even among Americans, American cars aren't considered that good. There's a massive reliability premium you pay for Honda and Toyota. Even cars with 100k miles on them (frustratingly as a buyer) keep their value. And they're manufactured in the US, inasmuch as any car can be said to be manufactured in a single location.
I've been searching around and I can't even find data about other countries importing our cars which to me would be the biggest signal of strength.
I own a Chevy Bolt EUV, made in the US. After 18 months driving it, I was happy enough with it that we leased a Chevy Equinox EV to replace my wife's gas car. The Equinox is made in Mexico, not the US but we've also been happy with it for 9 months so far.
We've owned Hondas (Odyssey) and Toyotas (Camry, Prius, Corolla). They've been great. We also changed the oil whenever the car's display said to and did whatever other servicing our independent mechanic advised. I suspect that a lot of cars would also be reliable if they were maintained.
Toyota is recalling 100,000 Tundra trucks because debris was left in the engine. https://www.haleytoyota.com/blog/the-2022-2023-toyota-tundra... There's no perfect vehicle although I'd say EVs get a lot closer when you can refill at home and do basically no maintenance except tire rotations and cabin air filters.
I have a 2018 Model 3 and your description of BYD is exactly how I would describe my Tesla. It feels cheap and plasticky and it creaks. I also briefly had a Model 3 rental car that was newer than mine (but I don't know what year it was) and it also felt the same.
I had thought about throwing an exception for Tesla because they did manage to create cars that people outside the US want. So I guess that does count but I doubt they're what anyone thinks of when they think of American car makes.
Oh I would for sure buy a BYD today if I were able. The ones I've ridden in have been really nice. I mean they are literally plastic but so is every car in the "economy" price range. I don't think their interiors were noticeably different than any other non-luxury car. I've been told that their higher end models don't have this problem.
The hyper competitive landscape only exists within China.
However, the much higher prices these companies are selling their cars outside of China are still much lower than the prices American cars are available at.
How many issues due large companies run into thinking they can just throw money at it? Just look at google and stadia, or amazon and their failed game studio. They have immense money and knowledge and ended up with nothing.
Each car has dozens to 100+ ecus, written in different languages, by different teams, different requirements, and different companies. Some are proprietary. Ford can’t just tell Bosch, hey your abs module needs to now integrate with our api, multiplied by 100+ companies. The legacy car makers need to revisit everything, and move most of it in-house.
At the same time, we've had car companies putting out cars for 20 years with 10s of different modules built by different companies and things have been working just fine. Suddenly it's a problem because apparently everyone needs a giant screen on the dashboard?
In the enthusiast market we're seeing the opposite. Evs have made speed routine, but dull the driving experience part. 90's and 2000's sports cars with manuals are worth sometimes 100ks more than those with paddels, even though they are slower.
The gtr has been essentially the same platform since its release, when the competition has gotten extremely tight. Then there's nissan's issues, ceo getting arrested, money troubles, uninspiring lineup that mostly sells to fleets/rentals, thus leading to inability to invest in halo vehicles.
If the plaid was making cars like this obsolete why are we seeing insane demand for the 911?
Whilst I understand the intention in comparing the average ICE to EV in order to get the larger upfront cost for the EV, I've never known anyone to shop like that. This argument was used against the Prius as well, "it's a 40k car where you could get something comparable for 25k, you'll never make up the cost of gas between the two". But that's not how the high majority of people shop for cars. They figure out their budget, then go around that. If you're in the market and can afford a 40k car, it's unlikely you'd go and get a 25k one, and vice versa, if you're looking for a 25k car, you're not going to go spend 40k because it's an EV.
So while it's nice to known that a more expensive EV will recoup its costs compared to a lower priced IC, for most, they're looking at a 40kIC vs a 40k EV, (post savings, rebates, etc) that hit the monthly payment they're looking for, and in that instance, the EV is going to win 99% of the time.
There’s a huge aftermarket for batteries and parts. Many companies will buy and refurbish battery packs as they are useful in many applications. A “dead” battery pack likely has many good cells and worth repairing.
When having my mini splits installed I pushed for an erv system in the bedroom. The installer had only ever done them in commercial units, and he hemmed and hawed about it, but I had two c02 monitors in my room showing it getting to above 2000 whilst sleeping. I've noticed a big improvement in how groggy I feel in the morning.
The flair 58 is 600+. So it's slightly more, but from a new manufacturer and has a unique design. I expected it to be at least a $1000, so I think it's definitely price competitive. Though I'd hold off until reviews come in.
So, the Flair 58 is their 'highest end' model and sports a 58mm portafilter... this thing is a 51mm portafilter which is basically only used on smaller portable machines. Also, it's listed on their website for $580. Not sure how you get to 600+ (unless you include tax, I guess).
A more comparable model from Flair would be either the Pro 3 ($325, all metal in the grouphead, pressure gauge, shot mirror... lots of included accessories) or the cheaper models they offer (Classic w/ pressure gauge, $230, Neo Flex, $99).
If you wanted to compare to the Cafelat Robot, that is also only $450... and is all metal, built like a tank, and has a very charming aesthetic.
The Meticulous is a lower end competition with the Decent Espresso DE1, or a more upscale version of the DIY stuff like Gaggiuino.
Given the author has given us essentially nothing regarding what is actually controllable (besides pressure control?), it's unclear to me what you even can do with it. A simple pressure control is pretty basic and not at all comparable to the Meticulous or a DE1.
- Group preheat (so it has some kind of heater)
- "Fully adjustable power" - ???
It does seem that the water tube either goes to a kettle and the pump is in the machine, or it goes to a pump, that you then need to attach to a machine, and that clear line is pressurized.
It does potentially have one feature the met doesn't (hinted at by allowing for filter brews): it'll be able to use up the entire water source, not a small amount of water you pour into the machine (similar to the Decent).
Edit: based on the manual just added, it seems like the pump is in the machine.
Best I can tell, with the manual released, this is literally just a pump and a group head. The dial appears to control the pump 'power' (voltage, I assume), and that's about it.
I'm really, really not seeing what could possibly justify this price. If the 'control' is just as simple as an analog knob, then this is no different than adding 'flow control' via a common dimmer switch to any other pump. I've done this modification on vibratory pump models myself, and they function just fine when dimer switch modded.
"Just a pump with a group head" is still pretty cool. It isn't revolutionary, temp control is a "big deal" in the community, but it's pretty cool that OP created that.
The pump absolutely is a "big deal" though if they can deliver on it. It has been attempted (Decent is trying to make one - actually they have been trying for years) and no one has delivered on a pump like that to date.
A small rotary pump would be innovative, nothing like that exists currently. Vibe pumps are loud, but they are tested tech and live forever. A drop in replacement rotary pump would have a big market if it can stand up to use.
Yes, they are pretty big. Just the pump alone is larger than the pump + motor for a vibe pump, and the motors for rotary pumps are like 4-5x the size of the pump. So you're really approaching 10x the size, it's kinda wild.
No offense, but we're not really comparing different things. You're offering an espresso machine that fundamentally has:
* No heating control
* No tank/water storage
* No milk frothing capability
The obvious comparison is a manual (lever) espresso machine that does not offer its own heating capability. It offers pressure control (via your arm) just fine.
Also, besides noise complaints and possibly some questionable reasoning involving vibe pump longevity, I have yet to see a compelling reason a rotary pump is better. They're 'nicer' and offered in higher end stuff, but performance wise a very good vibe pump seems just fine. Flow rates are more than adequate for pretty much any normal brewing method.
Regarding 51mm vs 58mm: you might be correct technically, but the ecosystem around accessories is firmly in the 58mm camp. As far as I can tell, the difference is so marginal it doesn't really matter anyways. Puck prep and other things will matter more for the average user.
I disagree. There's a very large espresso market for those who want a cheaper device that doesn't have the complications of a boiler. I have a flair espresso machine myself.
Adding a boiler I'd guess would double the price, so I think it's a good decision to leave off.
It’s not under a decade for ai to get to this stage but multiple decades of work, with algorithms finally able to take advantage of gpu hardware to massively excel.
There’s already feeling that growth has slowed, I’m not seeing the rise in performance at coding tasks that I saw over the past few years. I see no fundamental improvements that would suggest exponential growth or human level of performance.