lysecret

7 ms·

1.
by lysecret 7d ago
What’s the best way to get this hosted with eu residency?
2.
by lysecret 14d ago
Also just want to let my appreciation here for 3.7 it’s cheap super fast super reliable incredible at information parsing eu host able (important for us) and perfectly integrated into gcp. Great job google!
3.
by lysecret 3mo ago
I don’t get it we are paying at least hundreds or maybe thousands per month on ai costs. Just get a regular vm ?
4.
by lysecret 4mo ago
Sadly I think the correlation between this benchmark and performance is starting to break down imo. Still a legendary idea will be remembered and ingrained in the models forever haha
5.
by lysecret 4mo ago
Honestly happy he’s back at a foundation lab. He will have insane impact there. Of course one of the best educators in the world it’s a bit sad he gave up building and education tool.
6.
by lysecret 4mo ago
Ai in my mind is a new primitive of computing, like compute a db blob storage.
7.
Goblin Hunt(blog.leloew.com)
6 points by lysecret 4mo ago|0 comments
8.
by lysecret 4mo ago
There is a great recent episode of latent space about a similar topic it’s worth a watch even with the click baiti thumbnail and title https://youtu.be/9d899Ram9Bs?is=pQMoVmlWVsTNKfRK
9.
by lysecret 5mo ago
What’s the default context window? Seems extremely short.
10.
by lysecret 6mo ago
Main reason to me is that its layers on layer on top of the base LLM calls with not so much to show for it. Also a lot of native features (like for examples geminis native structured responses) aren't well supported.
11.
by lysecret 6mo ago
Because we are incapable of measuring developer productivity.
12.
by lysecret 7mo ago
Im honestly not that much worried there are some obvious problems (exfiltrate data labeled as sensitive, take actions that are costly, delete/change sensitive resources) if you have a properly compliant infrastructure all these actions
13.
by lysecret 7mo ago
Please I need 3 in ga…
14.
by lysecret 7mo ago
Local really has a strange meaning when most of what these things do is interact with the internet in an unrestricted way
15.
by lysecret 8mo ago
It’s just fake accounts being created the bot network has a bot problem haha
16.
by lysecret 8mo ago
There is son much personal info in here it’s wild.
17.
by lysecret 9mo ago
Hmm so that post also links back to segmentation done by structured outputs? (Though here not even enforcing the structure) https://ai.google.dev/gemini-api/docs/image-understanding#se...
18.
by lysecret 9mo ago
Oh I didn’t know about the visual bounding boxes this is super cool! Quick question are you talking about this feature? https://docs.cloud.google.com/vertex-ai/generative-ai/docs/b... Because it’s just using
19.
by lysecret 9mo ago
It’s funny how every podcaster/public ai figure is so certain text as a Ui will go away and it’s not going anywhere.
20.
by lysecret 9mo ago
3 flash is also insanely good even slightly outperforms 3 pro for me.
21.
by lysecret 9mo ago
Super excited about this generally ok satisfied with pyright but so I was with conda before uv or black before ruff.
22.
by lysecret 9mo ago
When I was debugging through f# code they definitely had that.
23.
by lysecret 9mo ago
I see this, a hot take form my side as someone who is bought in to GCP i quite like being able to put everything on the same billing account / handle it easily through service accounts.
24.
by lysecret 10mo ago
2.5 pro is already excellent at this.
25.
by lysecret 10mo ago
Now compare on free cash flow
26.
by lysecret 10mo ago
Yep fully agreed the main thing is to break apart the systems so any retries don’t lead to issues like you mentioned. I do still think there is sufficient amount of boilerplate to potentially justify some engine like this.
27.
by lysecret 10mo ago
Cursor has this too
28.
by lysecret 10mo ago
Yes very good point this to me is one of the most magical elements of this loss how it suddenly makes the model "collapse" on one output and the predictions become sharp.
29.
by lysecret 10mo ago
Yes you are right it's applied to the parameters, but other models (like ngcm) applied it to the inputs. IMO it shouldn't make a huge difference main point is you max differences between models.
30.
by lysecret 10mo ago
To encourage diversity between the different members in an ensemble. I think people are doing very similar things for MOE networks but im not that deep into that topic.

More ›