ubutler

4 ms·

1.
by ubutler 12d ago
If anything, the latest generation of AI models, Astra and Fable, are prime example of overfitting—whereas benchmarks suggest they’re AGI-tier, users (including myself) report the same old gaslighting, hallucination, context rot, cheating,
2.
by ubutler 2mo ago
Just a word of advice, it could go a long way to have a human write the README rather than an LLM. Right now, it’s a little hard to follow what the actual findings are, why they are what they are, and why that’s important.
3.
by ubutler 3mo ago
> The six solid objects discovered on Forrest Beach, to the north of Townsville, are thought to be space debris, and the Australian Space Agency (ASA) is now trying to determine where they came from. The BBC has approached the agency for
4.
A Source of Mysterious Repeating Radio Signals from Space Has Been Identified(wired.com)
4 points by ubutler 3mo ago|0 comments
5.
Google Can't Math Parsecs(lesswrong.com)
13 points by ubutler 3mo ago|4 comments
6.
by ubutler 4mo ago
> Sounds nice except that these are 1 very small scale model, 1 reranker, and 1 embedding model that are far from frontier LLM level. We've tried to take a first-principles approach to our end goal of 'legal superintelligence&#
7.
by ubutler 4mo ago
Our models can be deployed on premises as well, though that is more of a bespoke offering at the moment. We've also been fortunate enough to have trained most of our models on our own private infrastructure. Your question raises a broa
8.
Our response to the US ban on Fable 5 and Mythos 5(isaacus.com)
98 points by ubutler 4mo ago|21 comments
9.
The Blackstone Graph(isaacus.com)
2 points by ubutler 4mo ago|0 comments
10.
Ask HN: What are Stainless users doing now that Anthropic has killed it?
5 points by ubutler 4mo ago|3 comments
11.
by ubutler 6mo ago
ChatGPT already does this, albeit in limited circumstances, through the use of its sandbox environment. Asking GPT in thinking mode to, for example, count the number of “l”s in a long text may see it run a Python script to do so. There’s a
12.
Introducing AI chunking to semchunk(isaacus.com)
2 points by ubutler 6mo ago|0 comments
13.
Show HN: Isaacus – the legal AI research company(isaacus.com)
1 points by ubutler 7mo ago|0 comments
14.
by ubutler 7mo ago
This is called a lexical innovation ;). https://en.wikipedia.org/wiki/Lexical_innovation . We'd argue it makes a lot of sense to appropriate 'graphitization' as a term for a model designed to transform da
15.
by ubutler 7mo ago
> Also, you really want to tell people how to access it and what it costs. Or put up a "call for quote" if your market is large Enterprise budgets. Our pricing page can be found in our documentation here: https://doc
16.
by ubutler 7mo ago
FWIW we're planning on releasing a self-hostable version on AWS Marketplace quite soon followed by one on the Azure Marketplace. In both cases, deployments live entirely in your tenancy, are fully air-gapped (ie, they can't access
17.
Show HN: Kanon 2 Enricher – the first hierarchical graphitization model(isaacus.com)
10 points by ubutler 7mo ago|6 comments
18.
Popular text editor Notepad++ was hacked to drop malware(itnews.com.au)
1 points by ubutler 8mo ago|1 comments
19.
UTC with Smoothed Leap Seconds (UTC-SLS)(cl.cam.ac.uk)
2 points by ubutler 8mo ago|0 comments
20.
by ubutler 10mo ago
> Weirdly, the blog announcement completely omits the actual new context window size which is 400,000: https://platform.openai.com/docs/models/gpt-5.2 As @lopuhin points out, they already claimed that context w
21.
Spoofing her majesty in the 'Great Royal Phone Embarrassment' of 1995(1995blog.com)
2 points by ubutler 10mo ago|2 comments
22.
by ubutler 10mo ago
Here's a copy of the only known recording of a prank call to the Queen: https://www.youtube.com/watch?v=-YFFhc3XZDw
23.
by ubutler 10mo ago
After having read the article in its entirety, I’m still not sure what Cybersyn is…
24.
Australia's High Court Chief Justice says judges have become "human filters"(theguardian.com)
6 points by ubutler 10mo ago|0 comments
25.
Show HN: The Legal Embedding Benchmark (MLEB)(huggingface.co)
11 points by ubutler 11mo ago|0 comments
26.
Euro cops take down cybercrime network with 49M fake accounts(itnews.com.au)
155 points by ubutler 11mo ago|92 comments
27.
by ubutler 11mo ago
Personally, I like it. However, I like being able to comment and upvote more. At the same time, I'd be reluctant to say the least to hand over my login credentials. It could be quite cool to see this turned into a FOSS RES-style browse
28.
Show HN: We built the first comprehensive benchmark for legal retrieval(huggingface.co)
1 points by ubutler 11mo ago|0 comments
29.
by ubutler 1y ago
We were unfortunately disappointed to discover that, yes, Voyage, Cohere, and Jina all train on the data of their API customers by default. Voyage's terms say: > you grant Voyage AI (and its successors and assigns) a worldwide, irre
30.
Introducing the Massive Legal Embedding Benchmark (MLEB)(isaacus.com)
7 points by ubutler 1y ago|4 comments

More ›