planb

7 ms·

61.
by planb 1y ago
The tricky race conditions are the ones you often don't see in the debugger, because stopping one thread makes the behavior deterministic. But that aside, for webapps I feel it's way easier to just set a breakpoint and stop to se
62.
by planb 1y ago
I am in control of the vector database and the search index. I have no control over the different accessed data sources that don’t even allow to query access rights per resource (and just allow for can_access checks for a given user)
63.
by planb 1y ago
I’ll answer to this as a placeholder for all the „just do xyz“ replies: Searching the whole index and then filtering is possible, but infeasible for large indexes where a specific user only has access to a few docs. And for diverse data sou
64.
by planb 1y ago
That would surely be a possible way, but I don't want to block anything, I just want reasonable expectations and a basic understanding of the problem on all sides.
65.
by planb 1y ago
Yeah. If you have knowledge stored in a structured form like that, you don't need an AI...
66.
by planb 1y ago
I am assigned to develop a company internal chatbot that accesses confidential documents and I am having a really hard time communicating this problem to executives: As long as not ALL the data the agent hat access too is checked against th
67.
by planb 1y ago
This sounds even more like a machine than the Kuzma though.
68.
by planb 1y ago
Have you tried repeating this a few times in a fresh session and then modifying a few phrases and asking the question again (in a fresh context)? I have a strong feeling this is not repeatable.. Edit: I tried it and got different results: &
69.
by planb 1y ago
You‘re right, but that just reflects the structural sexism in our society while the wording by the op was intentional (I suppose. If not, I might as well be more sexist then he is).
70.
by planb 1y ago
Yes - people in HR departments are often female and often clueless, but I don't see the parent denying this. The wording of OP connected both though, which is sexist and can be considered "evil".
71.
by planb 1y ago
Seams weird to not have "How does this compare to Apple Intelligence Writing Tools" at least in the FAQ. Maybe refine is better or has more features, but the page doesn't even seem to acknowledge that a system level feature l
72.
by planb 1y ago
No, this is under the DSA (digital services act), not the DMA. The DSA tells platforms how they must keep users safe and transparent, the DMA tells the largest gatekeepers how they must behave toward competitors and business users.
73.
by planb 1y ago
Wow, I did not know that, thanks for making me aware of this. I guess the tech press only covers the big cases? I didn't even know there were so many cases already. edit: I looked it up - you are talking about the DSA (digital services
74.
by planb 1y ago
As a former European, I agree with your first statement. I love that the EU is taking this seriously, and I like how they introduced the "gatekeeper" term to apply regulations only to the "big ones" and not small busines
75.
by planb 1y ago
How can you ensure that your phone is not recording you 24/7? It’s called „trust“. Trust in the vendor, trust in independent 3rd parties that would identify such an issue, trust in capitalism because this would certainly be very bad pu
76.
by planb 1y ago
I think the time has come to not mean LLMs when talking about AI. An agent with web access can do so much more and hallucinates way less than "just" the model. We should start seeing the model as a building block of an AI system.
77.
by planb 1y ago
And by a sample that has become increasingly known as a benchmark. Newer training data will contain more articles like this one, which naturally improves the capabilities of an LLM to estimate what’s considered a good „pelican on a bike“.
78.
by planb 1y ago
> Besides a non-ad-free experience, the $7.99 Premium Lite subscription also removed several other cool features like the ability to download videos for offline viewing and background play, which can be quite convenient for listening to
79.
by planb 1y ago
Why? I’d guess requests from a local network site to itself (maybe even to others on the same network) will be allowed.
80.
by planb 1y ago
I don't know you or your work. But I know (senior) developers that don't use a debugger and I've often noticed that when fixing a bug, they don't build a full mental model of what really goes wrong and often just fix the
81.
by planb 1y ago
No the name of the user is part of the system prompt. How would you think this works? You can get ChatGPT to tell you all it knows about you, which is more than memories but never anything out of old conversations
82.
by planb 1y ago
No - but a random hacker news commenter wouldn't be the only one noticing this.
83.
by planb 1y ago
Sorry, but I refuse to believe you until you provide proof. What exactly did it remember? I think you are misreading the hallucinations here. If this was true, there might even be laws here in Europe that they are breaking.
84.
by planb 1y ago
This is not true. How do you come to this conclusion?
85.
by planb 1y ago
>Claude NEVER repeats or translates song lyrics and politely refuses any request regarding reproduction, repetition, sharing, or translation of song lyrics. Is there a story behind this?
86.
by planb 1y ago
> We’ve seen developers doing amazing things with Gemini 2.5 Pro, so we decided to release an updated version a couple of weeks early to get into developers hands sooner. Today we’re excited to release Gemini 2.5 Pro Preview (I/O ed
87.
by planb 1y ago
Because it‘s not a feature of the LLM but the product that is built around it (like ChatGPT).
88.
by planb 1y ago
What is wrong with OpenAI? The naming of their models seems like it is intentionally confusing - maybe to distract from lack of progress? Honestly, I have no idea which model to use for simply everyday tasks anymore.
89.
by planb 1y ago
Exactly what I thought when reading this. And that's the moment I stop taking the (probably somewhat valid) claims seriously. This reminds me of the arguments against streaming we had a few years ago, that confused compression/dec
90.
Justice Dept. scales back crypto cases in line with Trump administration memo(washingtonpost.com)
119 points by planb 1y ago|234 comments

More ›