Yeah they're very much deniable. Raw LOC/hr is much higher, and putting together a MVP, but I've yet to see any evidence that an LLM is capable of doing anything unsupervised, and if you need a human supervising everything it does... why bother having an LLM in the first place?
Because it can perform much faster? Monitoring allows you to multitask more effectively. I would also disagree that you can’t one shot anything…claims like this are weak and I have enough counter examples in my own life that it’s trivially false. The question is more: can it one shot the right things with a low enough failure rate for it to be a good replacement. It’s hard to figure that out a priori.
Yes, I bought it to make apps when it first came out, but I immediately picked up a new Shopify client and couldn't justify the time/investment. I had to choose between dunking $5k on it to maybe develop apps when my work schedule opened up, or return it and get the money back.
That AI segment was a boomer core slop fest. But to be fair, it's clear that Apple is not on the AI bleeding edge and it appears it doesn't want to be, it cannot afford to ignore it tho.
Let's hope they don't get overconfident with Gemini and pull a MS Copilot..
I get this vibe too. Turning Siri into yet another chatbot is a far cry from the vaporware they showed at 2024's WWDC. Seems they found out LLMs can't actually do that, but investors aren't just going to let them ignore it unfortunately.
Feels like they are just phoning it in here and waiting on AI hype bubble to burst. "Here's your stupid chatbot, now shut up"
Don't want to be too harsh, maybe I'm missing something, but the CPU is at least 2 years old, internally it has been a complete shitshow and that's a minor hiccup when compared to the firmware and software situation.
It's an interesting "newcomer" and the more the better but calling this a "beast" and a "game changer" is ridiculous to say the least.
I get the feeling this also means AI works very well for the general coding tasks and that's their biggest success in terms of difficulty AND people paying for it.
Of course every AI company has been over promising and pumping the numbers as much as possible but OpenAI has been hitting the reality wall more because both their people not being able to keep improving at a faster rate and their whole cost structure and financial plates spinning.
This doesn't invalidate the fact Anthropic is also overhyped to the max for their IPO.
It makes parts of the CIA look incompetent to the public. This is rare.
It’s reasonable to assume they knew it from the start he was getting money illicitly from the Navy and they might have enabled it. This was part of the leverage they had on him, to be used if he ever became a liability.
> It makes parts of the CIA look incompetent to the public. This is rare.
Iran–Contra? Their cyber espoinage tools getting swiped? The self-admission from the US administration that the CIA failed to properly recognize and prepare for the tactics used in 9/11, in spite of ample forward warning?
Like most intelligence agencies, their unaccountable power often gets mistaken for actual intelligence. That power makes them dangerous, but it doesn't make them smart.
They also totally missed India's second round of nuclear weapon development and were blindsided by the tests:
U.S. Intelligence and India's Nuclear Tests: Lessons Learned
August 11, 1998 98-672
The U.S. Intelligence Community did not have advance knowledge that India intended to conduct nuclear tests beginning on May 11, 1998.
Although intelligence agencies cannot have foreknowledge of every significant development in world affairs, many observers (and senior intelligence officials) believe that, in view of the election of an Indian government committed to "inducting" nuclear weapons, much greater attention should have been given to indications of impending nuclear tests
I reject the the idea that these types of people are needed. It's probably that most of the people in the CIA happen to be like that because they're power-hungry and they're just selecting their kin and justifying their choices as "right kind" because they narcissistically believe themselves to be the right type... They're probably the wrong type. Especially if they all share narcissistic or psychopathic traits; it's too many, it cannot work.
The company hires people who match the company’s desired culture.
If the people in the CIA who do hiring want the talent who are excellent at lying and compartmentalizing their ethics, then that’s what the organization becomes over a generation.
Ha, this was explored in some credible article here on HN some months ago, wasn't it? It made complete sense.
You can't be a nice balanced guy doing work which often dips in shady stuff, sometimes being responsible for killing innocents, or in extreme cases one's failed actions can send some region into death spiral of some small or larger conflict. No, you need (relatively) smart folks for whom emotions are just a tool to use on others to achieve your goals.
And this obviously has various side effects, some quite negative.
the CIA is literally tasked with breaking (other countries') laws. tradecraft is a very similar skillset to being an effective criminal.
think about it: shell companies, lockpicks, bribes, theft, blackmail, hacking, forgery. two kinds of people do those things: spooks, and the mob. the difference is why you're doing it and to whom.
also, if anything the CIA is far tamer today than it was in the '60s.
A lot of the really sketchy stuff the CIA used to do has been folded under special ops parts of the military. After the church committee, the CIA has to report to the senate intelligence committee. The military only really answers to the Commander in Chief(POTUS), and gets away with a lot more.
Check out the book "The Fort Bragg Cartel". Tl;Dr, the US military and special ops were holding up the poppy industry in Afghanistan, something like 80% of the world's supply of Heroin came out of US occupied Afghanistan. The DEA would look the other way on shipments intercepted over a certain size.
The special ops guys brought the drug trafficking home, now i95 through the southern states is a major drug trafficking route.
Got it, thanks. The CIA's cocaine was still ending up on the streets of America during Iran-Contra, though, so I (in my very uninformed opinion!) don't feel like the CIA cared that much.
It’s likely the cocaine originating from Colombia was going to end up in the US with or without the CIA’s help.
Also, it’s worth reading about the journalist who broke the story about Contra-sourced cocaine making it to LA. That journalist is very often misquoted and when he was still alive he tried to fix the missing several times.
All spies are bastards. That's sort of their job. In the CIA it might speak more ill of the guy who was arrested that he was arrested than that he (allegedly) inflated his credentials and might have bilked the military for leave pay.
When I worked at Los Alamos, the head of counterintelligence was a former CIA ops guy. He seemed warm and intelligent; just an all around good guy. In presentations, he explained that his job at the CIA was to exploit that appearance to ruin people's lives. He said his approach was to gain the trust of a mark, and then provide him with what ever was required to get him to betray things he loved. The presentations were effective, enlightening and spooky.
Yeah, that's why in a functioning State you have means to control the damage. But now we seem to have accepted it is a free for all and just throw ours helpless hands in the air and hope we are next to enjoy the criminal bonanza at some point.
Someone inside Ferrari had the terrible idea of greenlighting this and even more terrible lack of courage to not cancel this mistake because it was the baby turd of Jony Ive and Marc Newson.
Fortunately everyone will laugh and cringe, the usual car "journalists" will bite their tongues because they don't want to lose access, time will pass and it will be forgotten because Ferrari can afford to make these mistakes ( for now.. )
It reminds me of a rant that my friend sometimes goes on with regards to really low quality items, particularly about music...
someone wrote it, someone performed it, someone mixed it, someone approved it, someone developed marketing for it, someone helped get it on shelves, and then someone played it.
There were plenty of points along the way where the disaster could have been averted.
I don't understand the point of the rant. What disaster is having "bad music" out there? Is it stealing storage from "good music"? I understand this kind of rant for an iPhone, where a shitty decision brought along the chain of approval will impact million of people that are more or less stuck in the ecosystem. But music of all things? How do you even get in contact with "bad music"?
A lot of peple in this chain aren't paid to have a sense of ownership. They just do their job and their personal opinion of the work doesn't really matter.
Some of us care. Standing up and saying the product is crap leads to being asked to leave (fired). Or ends up on deaf ears, and the product is hated by people. Been in both situations, it doesn't seem there is a winning position.
I've been in the "someone performed it" and "someone mixed it" role for some tracks that I found utterly mediocre and yet ended up being some of the most successful stuff I've ever worked on. I mean, sure, previous works, marketing and hype can do a lot to alter the general perception, but most of the times it's just matter of being the right audience.
Missteps both in music and in other areas don't usually kill something that wasn't already moribund. The trashcan Mac Pro didn't kill Apple; Procol Harum's cover of Eight Days a Week didn't kill them or the Beatles.
They are priced for wider appeal and a different target group. At my local dealer I have the impression it's mostly a certain kind of owners (who got it from their partner that bought a 911) but that's purely anecdotal. Don't think this works for Ferrari, but then again I see also quite some Lamborghini Urus which I will never understand
"(for now)" is important, Jaguar used to have luxury-performance status by the neck - and they used their affordance of failed product luxury too excessively. Now, they're in a hole they cannot escape.
I think they have to make and sell some EV, just to have experience of it. If it isn't attractive, that doesn't matter. You can't, in this year, be so behind in EVs that you haven't ever sold one to customers if you are to be expected to make cars in the longer run, because in the medium term, even things like petrol stations are going to disappear.
The only complain I have ( not really directed at the article, but.. ) is to put all these theories and somewhat private experiments into the same room as pure gambling schemes turbocharged by "the algorithm" and political corruption.
While far from Heaven's gates, some guy trying to predict the price of corn next year is not in the same plane as those who had the "very original" idea every guy in his early 20s had at some point but never went further because he read some articles about "the law". Like it or not, the laws or the remnants of it were put in place due to the obvious degenerate attitudes and it's consequences gambling was always known for.
And no, it's not a "market", even Uber appears to have some usefulness to offset all the lying, corruption and criminality they had to do in order to become what they are. These ones don't even take you places other than gambler addiction.
Great to know, but what was the cost both in terms of $$ and tokens used?
Not to invalidate these benchmark results because they are useful, but the real usefulness it what they are capable to do when real people interact with them at scale.
Regardless, these are good news, because now that Microsoft is basically giving up their all-in strategy with Github's Copilot and Anthropic is playing the "I'm too good for you" game, it's about time for them to get pressed into not making this AI world into a divide between the haves and the have-nots.
You’d be surprised with some long running complex tasks. I’ve seen Kimi spend 8 minutes (total) thinking on a task that Claude got done in 30 seconds. They both ultimately got it right, but Kimi spent ~$2.25 to Claude’s ~$0.20
The jury is still out on that.
reply