I do think Apple needs a better Siri, but I think ultimately they were smart not to plow tons of money into it trying to do it themselves.
A better Siri is an expense to keep up the premium brand, not something that they will monetize. For particular uses of AI people will just want particular apps.
The memes are very funny, but I also am unsure what to make of it. Is the "unusually busy" feature mean like once a year business or one out of every ten days?
I am not bearish on AI by any means, but don't really get why Apple needs the best AI in-house other than to improve Siri. That doesn't take 500B.
As long as you can download apps and set permissions for communication between apps/devices, then it seems like 3rd party developers will make Apple devices more valuable for them.
I definitely read less than I used to. I think in the past some books didn't need to be the length they are. I still like narrative non-fiction and for those the level of detail is the point.
It is harder to get back into reading, but think if you force yourself a little at first your brain gets "rewired" back to it just fine.
>The provision was squeezed into the bill, nicknamed the “Big Beautiful Bill,” in May. It is designed to prohibit states from “[enforcing] any law or regulation regulating [AI] models, [AI] systems, or automated decision systems” for a decade.
I still am a bit confused by this. States require certain tasks be done by license, like writing prescriptions or having a degree to teach in public schools. It seems like licensing rules like that could still be passed?
Though still makes me nervous. I admit I would like some type of safety inspection of the top labs, even though that sounds heavy handed. I just feel like it would be good for making them dedicate a minimum amount of resources to safety.
I think this is mostly right, but also I'm not sure I agree completely with the premise. Humans have years of conversations they've heard before they attempt to read or write. They already have a concept of what a 'dog' is before they see the word, and know what it is likely to do. Not the same with something that only sees text.
If anything I imagine law and cases to get more complex. Less lawyers may be needed for any one case, but that should just make it more possible for lawsuits for smaller amounts of damages to be brought.
The blog post brings up the Daniel Penny trial. That trial took six weeks, even though it was on video. I imagine criminal cases have gotten longer and more complex now, because each case has more potential information than it ever did.
It is interesting. My two theories would be a) people at top schools follow these meta-discussions more and are therefore are more likely to be ahead of trend b) someone who could do comp-sci at Stanford could easily become something like a doctor, but at small schools there's not an obvious second option that's also high earning.
I've long thought that if AI can start solving unsolved math problems we are in for a very weird (or bad) future, since at that point it really can come up with things that no one has thought of before. And perhaps, it would also be able to find new physics or new physical mechanisms very fast.
That said, it sounds like teams are guiding the areas of exploration pretty directly. If I was more cynical, I'd say that Google has a lot of financial incentive to claim it was integral to any breakthroughs the team comes up with.
I'm not anti self-driving cars in general, but I do think as they get more popular they should be taxed extra for every mile they drive without a person (or at least without cargo).
A self-driving car may decide to drive around an empty street in a city instead of find parking, because it's cheaper, but its a negative externality.
Yeah. For understandable reasons that is covered a lot too, but AI 2027 is really about the risk of self-replicating AI. Is an AI virus possible, and could it be easily stopped by humans and our military?
I do agree generally with this, but AI 2027 and other writings have moved my concern from 0% to 10%.
I know I sound crazy writing it out, but many of the really bad scenarios don't require consciousness or anything like that. It just requires they be self-replicating and the ability to operate without humans shutting them off.
There is an element that when it becomes cheaper/easier it cancels each other out. Spam became cheaper/easier but that just meant people were less trusting of random emails.
It probably is true that some populations are more vulnerable to AI produced propaganda/slop, but ultimately I have 50 more pressing AI concerns.
>For their study, the team assigned three groups of participants – students 18 to 39 years old from five universities in the greater Boston area – to write an essay, with the first group tasked to leverage LLM – ChatGPT, to be precise – to complete the task, the second to utilize search engines, and the last one to rely exclusively on their brains.
The headline is framed as people using ChatGPT as the treatment group, but in reality I feel like writing an essay with ChatGPT is the control group and people forced to write an essay is the treatment group. I think the headline should be "writing essays makes you smarter". Interesting conclusion, but also most people never write essays after college so I'm not exactly sure what the takeaway for society is right now.
(I still have my skepticism about the brain scan conclusions, but not my area of expertise)
It would be trivially easy to build an algo that beats the market 90% of years but 10% has huge losses. Just construct a slightly levered portfolio, with some protection against minor losses, but totally exposed to big losses.
If you sold that, it would appear like its working for awhile and you'd probably have lots of trusted customers by the time it fails.
(I'm a bit cynical, I worked in finance briefly, and I realized the fund we were selling to investors was essentially this)
In college (about 15 years ago) I worked for a professor who was compiling precint level results for old elections. My job was just to request the info and then do manual data entry. It was abysmally slow.
This application seems very good - but still a bit amazing that lawmakers haven't just required that all data be uploaded via csv! Even if every csv was slightly different format, it would be way easier for everyone (LLM or not).
I wouldn't touch Uber stock, but the PE ratio is only 15. My guess is the market is expecting them to be able to still grow in the next few years even if they eventually face tough competition.
I think there's also the fact that if self-driving cars take off and price goes down, people will ultimately rely on taxis/delivery more than ever. Maybe there is a place for Uber to be the platform for that still, maybe not.
Interesting Q - though I predict they will largely look the same. They may have more battery charging stations or replacement batteries (especially for stations on an interstate) but otherwise I think it will just be a nicer version of what we have today. Maybe higher quality hot food and more interesting snacks, which I think we are already seeing today with new stations.
Most gas stations today only have one employee at a time so it seems hard to cut that down (and not much incentive tbh).
Electric cars today are still under 10%. Cars last a really long time, so even if there is a huge surge in electric self-driving cars by 2030 there will still be lots of ICE cars on the road in 15 years.
A better Siri is an expense to keep up the premium brand, not something that they will monetize. For particular uses of AI people will just want particular apps.