Well not really. Apple has never really been first to market, but for the first decade of iPhone they never felt behind. Now the iPhone just feels behind in nearly every way but the SoC and videography. Battery, display, charging speeds, software, multitasking, AI, virtual assistant. I still enjoy using my 15 Pro Max, but the hardware and software of say the S25 Ultra makes the 15 Pro Max feel 5 years old.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Report Reveals Internal Chaos Behind Apple's Siri Failure
- Thread starter MacRumors
- Start date
- Sort by reaction score
OMG! What is the point of looking up general knowledge questions, if there is only a 50% chance the answer going right!?!
With those rates, just don’t bother learning whatever it is you think you are learning.
I use it for both, although get more "bang for my buck" for work uses, obviously. I treat it like an intern or an assistant I don't know well - if anything is mission critical it gets double checked, but honestly it doesn't screw up as much as you would think if you just read the news and see things people post online. Professionally I lead proposal development for my company, so I do a lot of analyzing large PDFs to figure out if something is worth bidding on, and if it is, assembling a team to write the proposal. (Note that at work I use a paid AI model that does not train on corporate data.) The proposal analysis literally saves my company tens of thousands of dollars a year, we were previously outsourcing that to an outside contractor because it's so time consuming.
Professional uses cases (just this week):
- Review a massive PDF request for proposal (RFP) to pull out contractually required deliverables and their due dates and build a schedule of deliverables in Excel (this would have taken me a couple of hours, it did it in a minute or two).
- Turn above schedule of deliverables into Gantt chart
- Determining which staff resumes on file best line up to position descriptions in RFP; tailoring them to address requirements
- Tailoring past performance narratives to better align to the requirements in RFP
- Perform competitive analysis to flag small business competitors or potential partners we may not have heard of
- Perform research on potential clients, review their social media feeds, news releases, etc. to pull out major initiatives and identify key themes that might appeal to them
- Grading proposal content against published evaluation factors; provided recommended areas for improvement
- "One voicing" proposal content (If you have four people writing different parts of a proposal you want it to sound like one person wrote it - things like making sure acronyms are only spelled out once, consistent use of Oxford comma, etc.)
- Identifying weaknesses or risks in an approach to do work so I can proactively address them in the proposal
- I take a picture of all my handwritten meeting notes and use AI to transcribe them digitally in seconds
- Performing company research on companies with openings
- Reviewing all copies of my resume (folder with 20+ years worth) and developing customized resumes to particular job postings
- Providing drafts of Cover Letters for those job postings
Some personal use cases over the past few weeks:
- Calculate the likely price of something after factoring in tariffs
- Translated an Italian ancestor's handwritten birth record from 1859 into English
- Reviewing my Mac's crash logs to determine what was causing Finder to crash repeatedly (it was TextExpander)
- Identify some activities that might appeal to my kid for a potential trip to visit my brother in Germany
- Identifying the cause of and fixing a weird error message I was getting on my home server
- Calculate calories and macronutrients in a meal based off of list of ingredients for use in meal tracking
- Develop six-week exercise routine to reach specific goal
- Summarized podcast transcript and flagged when conversation I cared about started
- Gathered numerous reviews of a product I was interested in and highlighting key takeaways
Thank You, appreciate you taking the time and the details!.
I can see the benefit you describe around RFPs etc having been involved in RFPs/SOWs etc myself.
Personal, not so much but part of that can be attributed that I grew up way before the internet and I travelled to ~ 40 countries without a (smart)phone or GPS or anything like that.
A new report from The Information today reveals much of the internal turmoil behind Apple Intelligence's revamped version of Siri.
Apple apparently weighed up multiple options for the backend of Apple Intelligence. One initial idea was to build both small and large language models, dubbed "Mini Mouse" and "Mighty Mouse," to run locally on iPhones and in the cloud, respectively. Siri's leadership then decided to go in a different direction and build a single large language model to handle all requests via the cloud, before a series of further technical pivots. The indecision and repeated changes in direction reportedly frustrated engineers and prompted some members of staff to leave Apple.
In addition to Apple's deeply ingrained stance on privacy, conflicting personalities within Apple contributed to the problems. More than half a dozen former employees who worked in Apple's AI and machine-learning group told The Information that poor leadership is to blame for its problems with execution, citing an overly relaxed culture, as well as a lack of ambition and appetite for taking risks when designing future versions of Siri.
Apple's AI/ML group has been dubbed "AIMLess" internally, while employees are said to refer to Siri as a "hot potato" that is continually passed between different teams with no significant improvements. There were also conflicts about higher pay, faster promotions, longer vacations, and shorter days for colleagues in the AI group.
Apple AI chief John Giannandrea was apparently confident he could fix Siri with the right training data and better web-scraping for answers to general knowledge questions. Senior leaders didn't respond with a sense of urgency to the debut of ChatGPT in 2022; Giannandrea told employees that he didn't believe chatbots like ChatGPT added much value for users.
In 2023, Apple managers told engineers that they were forbidden from including models from other companies in final Apple products and could only use them to benchmark against their own models, but Apple's own models "didn't perform nearly as well as OpenAI's technology."
Meanwhile, Siri leader Robby Walker focused on "small wins" such as reducing wait times for Siri responses. One of Walker's pet projects was removing the "hey" from the "hey Siri" voice command used to invoke the assistant, which took over two years to achieve. He also shot down an effort from a team of engineers to use LLMs to give Siri more emotional sensitivity so it could detect and give appropriate responses to users in distress.
Apple started a project codenamed "Link" to develop voice commands to control apps and complete tasks for the Vision Pro, with plans to allow users to navigate the web and resize windows with voice alone, as well as support commands from multiple people in a shared virtual space to collaborate. Most of these features were dropped because of the Siri team's inability to achieve them.
The report claims that the demo of Apple Intelligence's most impressive features at WWDC 2024, such as where Siri accesses a user's emails to find real-time flight data and provides a reminder about lunch plans using messages and plots a route in maps, was effectively fictitious. The demo apparently came as a surprise to members of the Siri team, who had never seen working versions of the capabilities.
The only feature from the WWDC demonstration that was activated on test devices was Apple Intelligence's pulsing, colorful ribbon around the edge of the display. The decision to showcase an artificial demonstration was a major departure from Apple's past behavior, where it would only show features and products at its events that were already working on test devices and that its marketing team had approved to ensure they could be released on schedule.
Some Apple employees are said to be optimistic that Craig Federighi and Mike Rockwell can turn Siri around. Federighi has apparently instructed Siri engineers to do "whatever it takes to build the best AI features," even if that means using open-source models from other companies in its software products as opposed to Apple's own models.
For more details on Apple's Siri debacle, see The Information's full report.
Article Link: Report Reveals Internal Chaos Behind Apple's Siri Failure
Apple AI chief John Giannandrea HAD GOT TO GO!
I've been saying this for months and longer! Literary hired cause he believed inuser privacy for his reason to leaving Google (and possibly the ickyfirst name familiarity of a version of "John") .
Anyone remember the LAST company that used end user privacy as a big part of their product marketing?! Hint their product was a fruit nake before changing their corporate name to match! See how well both those actions went for them?!
John is Google's Trojan horse into Apple not dissimilar to Microsoft did with an executive becoming the CEO at Nokia!
"More than half a dozen former employees who worked in Apple's AI and machine-learning group told The Information that poor leadership is to blame for its problems with execution, citing an overly relaxed culture, as well as a lack of ambition and appetite for taking risks when designing future versions of Siri."
Forstall wasn't let go for refusing a public apology for vector based data from TomTom he and his team had NO fault over. He MADE Apple money. While this guy John is costing Apple money??!
Tim let Johngo immediately: keeping him anchoring him makes nodollars ans nosense!
DJ Quick - Dollars & Sense
"IF it don't make dollarz it'd Ontario makes sense"!
Maybe they should look up some of the General Magic guys. This sounds like the conversation I had with Andy Hertzfeld while interviewing with the company in 1995. He was animated and enthusiastic about the idea of a cell phone that could be one's personal agent (of course, based on Magic Cap).
I am sure that it has already been said, but this was so dumb for Apple to do. Why would you make vast claims about abilities which you could not achieve. This Elizabeth Holmes style behaviour is extremely worrying for Apple. Apple should admit it was too ambitious. Instead it seems like Apple wants to smash down the wall and allow all sorts into the system which makes them Apple.
I think people are surprised because generally Apple delivers on promises. So we would assume that internally Apple was further along than they actually were.
Last edited:
I wouldn't be surprised if 2-3 years down the road, when Apple finally has a usable set of AI models and strategy, we are required to have a minimum of 32 GB RAM in any Apple device to run it. Like many other tech companies in history (remember Bill Gates saying 640 kB is all you'll need?), Apple got caught trying to flex its hardware superiority only to find in the market that it wasn't that superior in this context. In the end that's fine. The industry will learn what it takes, the players will adapt or die, etc. Eventually everyone's going to want to perform a decent percentage of this work on device because it will be more efficient than the wasteful distributed model currently in use. I'm still waiting for a killer app for me because so far I'm not that impressed by AI as a practical technology.
Last edited:
I have wondered the same about base memory, and have wondered if some of the issues we have now with AI, or some form of it on the iPhone's is due to memory usage. Now take into account the number of future apps that will be offering AI...gonna need lots of memory.
Looks like the top got too comfortable with constant selling figures and that relaxed attitude dribbled down to the rest of the company
He couldn't have been more wrong.
ChatGPT was the world’s most downloaded app last month - 9to5Mac
ChatGPT was the most downloaded app in the world last month, per an Appfigures report, and Studio Ghibli no doubt deserves a lot of credit.
9to5mac.com
"Senior leaders didn't respond"? You couldn’t get more reactive than that. It’s a terrible sign that Apple has lost its way. Sorry to say, but there are no true thought leaders left at Apple. If Steve were still around, Siri would’ve evolved into a powerful AI long before OpenAI even thought of GPT.
Their fear of making mistakes is the issue here. DeepSeek and Grok both show how quickly they could’ve trained an effective on-device/small memory and off-device/cloud LLM, respectively. Yet because they’re afraid of bad press, and there’s no visionary leader where the buck stops, they’ve sat on their hands and are now farming out reduced capabilities to OpenAI and Google.
I’ll echo your company research one. I used a free API that interfaces with Google Sheets to query OpenAI for a list of companies matching certain criteria and output contact info and intro messages to send to recruiters.
Mind sharing which you’re using for the resume stuff? Especially having it scan your resume folder and write cover letters? I’m curious whether it’s monitoring a folder and you can ask it to generate a resume or cover letter at any time, or you have to attach the folder each time.
I’ll echo your company research one. I used a free API that interfaces with Google Sheets to query OpenAI for a list of companies matching certain criteria and output contact info and intro messages to send to recruiters.
Last edited:
I’ve done it with both ChatGPT and Claude. If you upload the files to a “project” then you don’t have to attach a resume every time, but I tend to make a new project and reupload for each application (I don’t change jobs often so I don’t have that many resumes), unless it’s at the same company or a very similar role. By keeping the projects separate by company, you can start the project with doing research on the company, and then generate the resume/cover letter, which in my experience, tends to result in better cover letters and resumes. I’m not aware of any tool that will monitor a folder, but that’s a good idea.
You do have to proof the documents it produces. I haven’t seen it outright lie, but I have seen it stretch my experience in a way that I’m not comfortable with.
As a made-up example, say I have been a project manager for ten years, and last year I managed one project implementing AI into company operations. If I’m applying for a job at OpenAI, ChatGPT might produce a line saying something like “Ten years’ experience managing AI implementation projects” - whereas I’d say “Ten years’ experience managing technology projects, including a successful implementation of ChatGPT into company operations”.
Vision Pro will be a massive hit once its size and weight issues are sorted. You’ll see Jarvis/Cortana in your field of view when needed and talk to them otherwise, seeing all your apps, and all your AR enhancements to life.
Vision is visionary but was released too early. Let’s hope we see some real AI improvements at WWDC!
Siri in the car is hopeless.
Siri: I can’t show you while driving.
Me: Ok. So read it to me like ChatGPT does!
Siri: Sorry I don’t understand.
Me: That’s the problem, Siri. You don’t.