doubtfuluser

7 ms·

61.
by doubtfuluser 3y ago
This shouldn’t be too difficult, the api allows writing such a tool and with the „assistant API“ it should hopefully be able to put attention on the right parts… So: git clone -> system prompt -> add files of repo to messages -> ge
62.
by doubtfuluser 3y ago
With the assistant API, am I wrong or is it now much cheaper to actually use the API instead of the web Interface? $20 would cover a lot of interactions with the API, and since it’s now also doing truncation / history augmentation the
63.
by doubtfuluser 3y ago
Didn’t have many interactions which BCG so far but in both we had, I was surprised at how much money they get for reshuffling information from what is all common knowledge and available in the net. I can see that this is something LLMs can
64.
by doubtfuluser 3y ago
Well, I would call a 7B and 13B model a large language model… but I appreciate your humor.
65.
LeoLM: German-Language LLM Research(laion.ai)
105 points by doubtfuluser 3y ago|84 comments
66.
by doubtfuluser 3y ago
I wouldn’t call this a „paper“. They are pretty silent on a lot of technical details.
67.
by doubtfuluser 3y ago
So what will this mean in relation to their recent Alexa announcements? are they using Anthropic in this announcement? Is it so difficult for a big company to build their own?
68.
Kids create fake nudes of girls in their class in Spain(businessinsider.com)
2 points by doubtfuluser 3y ago|4 comments
69.
by doubtfuluser 3y ago
Im afraid this is just the beginning of what we are going to see in the near future. We are not prepared for this.
70.
by doubtfuluser 3y ago
Is there any information on the resolution of the outputs? This is indeed a new low in announcing something: „hey we have this awesome new model, which right now none can use, and we won’t tell you what is changed - except for it’s magicall
71.
Amazon announces new generative AI version of Alexa(venturebeat.com)
3 points by doubtfuluser 3y ago|1 comments
72.
by doubtfuluser 3y ago
Interesting. Looking forward to test this and see how good it is.
73.
by doubtfuluser 3y ago
That’s a very common theme at Amazon, afaik called two way doors. Identify if a thing is a two way door and if so, trying it out doesn’t need as much deep dives and safety as a one way door since it can easily be reversed. I think that’s a
74.
by doubtfuluser 3y ago
On „General Magic“, recently read „Build“ by Tony Fadell. I can recommend that for people interested in building startups. Will also give a bit more background on General Magic
75.
by doubtfuluser 3y ago
For me all of the current voice assistants have issues. Playing music seems to be super hard to get right if you don’t follow mainstream. Either you get a playlist with always the same titles, or it starts playing some obscure remixes. Home
76.
by doubtfuluser 3y ago
A bit of an unpopular opinion as it seems, but I would actually bet that the current prompt engineering is just a short term thing. When the performance of LLMs continue to improve I actually expect that they will become much better to unde
77.
by doubtfuluser 3y ago
This is really a fun project! Is it actually possible to use the code and create an app for sale out of it. The MIT license says so I think, just wondering if anything else would speak against it.
78.
Letting a ChatGPT with pinecone run infinitely to solve complex tasks(twitter.com)
5 points by doubtfuluser 3y ago|0 comments
79.
by doubtfuluser 3y ago
[Edited] Isn’t the copyright question a red-Hering? We are talking about models on the verge of generating output not distinguishable from human output. How is then a copyright breach - if it’s only caused by AI, but not by human - enforc
80.
by doubtfuluser 4y ago
I think it’s going to be a great enabler. A lot of people who now work for companies will be able to make use of the new tools to build their own companies or products. Suddenly one experienced developer can move more quickly and even get o
81.
by doubtfuluser 4y ago
So amazing! Today they would have mounted the grip in the other direction… it’s very difficult to make selfies with this grip…
82.
by doubtfuluser 4y ago
I think this is the curse but also the opportunity of AI. OP can now do the same work they did before in just 2-3 days. Could they now open their own studio and offer services for many customers on their own instead of working for one compa
83.
by doubtfuluser 4y ago
A whole new meaning comes to adversarial attacks in Neural Networks…
84.
by doubtfuluser 4y ago
As long as we have articles claiming language xyz replaces C++, c++ will be fine. The day none writes about c++ anymore is when it will be replaced.
85.
by doubtfuluser 4y ago
I bet Marc is also giving himself a „meets most“ and putting himself on a pip. Since as we know „he takes the responsibility“ for what is happening…
86.
by doubtfuluser 4y ago
Pretty funny to me that in the article it speaks about “the WhatsApp owner” almost as if there are now enough people who might not really know Facebook anymore…
87.
by doubtfuluser 4y ago
+100 on this article. I observed the same in two stages of my life. 1. When I became more senior scientist - scientists usually write crappy code - but once I had a lot of other stuff like coaching, strategic planning etc on my plate I was
88.
by doubtfuluser 4y ago
All the projects they mention going on after the wind-down are US centric. Will they do anything outside of the US, too?
89.
by doubtfuluser 4y ago
This is cool! You can actually get a hint for Kubrick‘s use of symmetry in his movies (except for „Dr. Strangelove“) Very nice!
90.
Amazon reviewing possible layoffs in Alexa and other businesses(businessinsider.com)
4 points by doubtfuluser 4y ago|0 comments

More ›