Tongue half in cheek, my point is that agents may be unreliable with fresh libraries.
It looks like there are some fun edge cases too:
> A reserved region can be active or inactive. For example, a reserved region that represents the fold is active when iPhone Duo is partially open, but inactive when it is fully open.
Note: I'm scrubbing it against the final 27.1 beta SDK as I type this, so it's likely that its Duo knowledge will get an update within the next few hours.
Fun fact, I’ve now worked on the surface duo (technically the windows version not the android version) and the iPhone duo.
I thought the ideas of the surface duo were great and I’m super excited to see them evolved and shipped in a product that I think a lot of people are going to love.
I think this is a good chance to try and port Alyx (ARM64 build) to visionOS and run it on my Vision Pro headset which has been running around unused for the last two years xD
> What modern LLMs do is apply brute-force computation to any domain expressible in language--not just a restricted chess domain. That's the algorithm.
Which means that companies with sufficient computational resources and money will be capable of unlocking problems thousands of times faster and more effective than any individual even when lacking the skills, just by a matter of try and error.
What's the value of using one of these phones running iOS? iOS is great for media consumption and not so much for productivity... so I don't understand very well the use cases... apart from watching Netflix I guess?
Most people aren’t doing productivity. They’re watching videos and playing games. This would mean I don’t need my iPad on flights or when travelling. And when I’m travelling I don’t need to be doing productive things.
iOS runs plenty of productivity software, many which run great on larger screen devices like this or iPads. What does "productivity" mean to you? Or is it just terminals and coding?
I might be a weirdo, but even the mail client on iOS has serious deficiencies. Just getting an email with a zip attachment where I have to edit / sign a document with a certificate / something, rezip and send back is a hell in iOS, while it simply feels natural in Android.
They had other models a while ago, named 'Pulsar', and they were finetunes of Nemotron in collaboration with NVIDIA. I think this will be something similar.
I mean, I think it depends. At home 3 of us we use AI for multiple reasons, from coding apps to asking general questions, and if we would have to pay equivalent subscriptions that would be ~1k a year on AI + submitting all your data to external services. I payed around ~8k on 2 DGX Sparks that, at the moment, serves perfectly fine as a ChatGPT/Claude replacement at home (DS4 Flash peaking at ~170 tokens per sec with 6 concurrent sequences), and even once the technology is obsolete for inference in a few years, I will still have 2 pretty powerful machines for whatever I need + some pretty fast NVME Storage. I don't think its a terribly bad idea.
I indexed thousands of documents into a SQLite database with an FTS5 index, plugged it into DeepSeek v4 Flash and got better much better results than any other commercial solutions my company has tried in the past.
The trick was just to let the LLM come up with its own SQL queries for searching... and the results are impressive.
Just wondering on the reported accuracy of the SQL queries generated by SQL based on RAG. Reports have been not encouraging and also confirmed by these results but perhaps your methodology is different from these reports [1],[2],[3].
>On this benchmark, a pure LLM generated an accuracy score of zero. Adding RAG, prompt engineering, and agentic AI raised accuracy to the 10+% range.
[1] Any text-to-SQL benchmark should address difficulties of real-world data stores (acm.org) (21 comments):
I'm now using this approach too, and it feels better than any sorting method I've used before. The only thing I'm thinking about now is: if the LLM makes a mistake, how can I provide feedback and verify it?
I log all toolcalls to a file, I think others have said the same. But I'm a bit leery of letting a hallucinating LLM write SQL queries. I think most I've spoken with, agree that an LLM is like a 20 year old, eager intern. Well meaning, but left unrestrained capable of immensely inexperienced mistakes.
Before a lot of frameworks existed, you'd see DEVs taking user input on a web form, and then just throwing it directly at the MTA. So spammers could submit email@address\nCC: persontospam@address, and the like.
Now LLMs are a different beast, but you have input validation for LLMs, unique to all other validation methods. Yet there's actually no safe way to ever validate user input for a LLM, except for very rigid input validation on single words. Take the email example above. You'd need a regex to only validate an email address (and that isn't simple), but once you expand it to actually allowing sentences?
The LLM is now input validation vulnerable.
And that means no user input can be used in unvalidated commands.
And then just random hallucinations. I'm curious how the gp managed weirdo LLM behaviour, like out of the blue 'drop table' or accidental select into as opposed to just select.
Its a read-only SQLite file. And I mean the people using the chatbot knows it uses AI so just like Google they shouldnt pick the first result, but forcing the LLM to mention the sources and not assume acronyms works amazingly well
I have a web user interface connected to a coding agent (OMP) running inside a container, so if any of the tool calls fail or something happens, usually my model recovers autonomously from these situations. The capabilities of models like DS4 Flash are those of frontier models from months ago, so its recovery and autonomous capabilities are quite impressive.
Not really, I just made a really simple Python library the agent can use and modify as he wishes within a sandboxed environment to explore the SQLite database. When the user prompts the agent, the prompt goes into my coding agent, omp, and starts running Python scripts and throwing SQL sentences until coming up with the answer. It works extremely well in our experience.
I am surprised. I've been using DS4 Flash (0731) for weeks now and it works perfectly fine as a replacement for Claude in a large variety of cases. It requires a few more iterations, sure, but it's useful enough to not need a Claude subscription anymore. Among the things I do I've been reverse engineering, writing complex C++ code...
DS4 Flash absolutely kicks ass for reverse engineering and bug hunting. Almost no point in considering paying for a bigger model, although it's possible the stuff I've fed it (wide variety of older DOS/Windows stuff and device firmwares) might be easier targets.
I've been reverse engineering LEON3-FT SPARC v8 BE code, so I wouldn't say it's common :D. When attached to Ghidra through a MCP the things you can do with this are simply crazy.
It depends, first I ask myself if this didn't exist would I vibe code it myself, if I would, then is this good enough, might as well save my time and tokens on checking than building from scratch. Personal projects wise
Curious what MCP setup you use? I'm not sure which one I have wired up, but I have to restart Ghidra every time I change file. I think it's either LaurieWired's original or a fork of it
Sure if you use remote AI services, but any companies working on niche markets where they want to protect their IP, or they simply work with sensitive stuff, will rely on local AI instead.
There’s no company that has “no sensitive stuff”, from HR to financials to customer data to board presentations to engineering IP and they all, without exception , entrust their data to cloud services for at least a decade now. Even governments do that, although they sometimes use special regions.
reply