mgreg

7 ms·

1.
OpenAI testing ads in ChatGPT free and Go tiers(twitter.com)
6 points by mgreg 8mo ago|1 comments
2.
by mgreg 1y ago
Link to repo for those interested: https://github.com/ZhangYiqun018/AvengersPro
3.
by mgreg 1y ago
There's also Rancher Desktop ( https://rancherdesktop.io/ ). Supports moby and containerd; also optionally runs kubernetes.
4.
by mgreg 3y ago
This is quite the loss for OpenAI. And no one seems to have heard what Ilya Sutskever's status is at OpenAI as well.
5.
Chinese Startup Is Winning the Open Source AI Race(wired.com)
5 points by mgreg 3y ago|0 comments
6.
by mgreg 3y ago
Unsurprising but disappointing none-the-less. Let’s just try to learn from it. It’s popular in the AI space to claim altruism and openness; OpenAI, Anthropic and xAI (the new Musk one) all have a funky governance structure because they wan
7.
by mgreg 3y ago
I think this is built into Excel already. In the Excel menu bar: Data -> Data from Picture
8.
by mgreg 3y ago
It's a great observation. People simply want their free stuff. The potential challenge arises in the future. Today's models will probably look weak compared to models we'll have in 1, 3 or 10 years which means that today&#x
9.
by mgreg 3y ago
Applying the term "open source" to AI models is a bit more nuanced than to software. Many consider reproducibility the bar to get over to earn the label "open source." For an AI model that means the model itself, the da
10.
If you can't reproduce the model then it's not open-source(twitter.com)
230 points by mgreg 3y ago|110 comments
11.
by mgreg 3y ago
This does look like a truly open model with all the components needed to replicate under Apache 2. This seems to be a fine-tuned version of their CrystalCoder model. Kudos for releasing a fully open model that will (hopefully) foster colla
12.
by mgreg 3y ago
I understand where you're coming from but what they provided DOES make their work reproducible. You can use the data, source code, and recipe to train the model and get the weights. It would be nice if they provided the weights so it
13.
by mgreg 3y ago
I very much appreciate that the authors not only published their code ( https://github.com/llm-random/llm-random ) but included the dataset they used (available on Huggingface - https://huggingface.co/dat
14.
by mgreg 3y ago
There are now several open source models that are fine tuned for function calling including: * Functionary [ https://github.com/MeetKai/functionary ] * NexusRaven [ https://github.com/nexusflowai/Nexu
15.
Andrew Ng: 'Do we think the world is better off with more or less intelligence?'(ft.com)
18 points by mgreg 3y ago|10 comments
16.
by mgreg 3y ago
The online models to a decent job of proving up-to-date info. Simple inputs like "who won the football game last night" provided the correct score and a little detail on the NFL's Monday Night game. Did well with some other
17.
by mgreg 3y ago
Some details that might interest you from SemiAnalysis [1] just published yesterday. There's quite a bit that goes into optimizing inference with lots of dials to turn. One thing that does seem to have a large impact is batch size wh
18.
by mgreg 3y ago
Agreed on 1; people in the court case are saying Google is better. Given that, why wouldn't Google negotiate a lower rev share? Consider this from Apple's Eddie Cue: "Cue implied that Bing’s technology was inferior to Google’
19.
by mgreg 3y ago
It seems unclear if Microsoft was willing to pay more. From Bloomberg's coverage: "Microsoft business development executive Jon Tinter said that his company weighed making a multibillion-dollar investment in its relationship with
20.
by mgreg 3y ago
There’s been some good points raised by both sides in the case but there’s still some logic that doesn’t flow. * Google pays a rev share to Apple for being _default_ search provided and default was key: “no default, no deal”. Apple turned
21.
Google User Data Has Become a Favorite Police Shortcut(bloomberg.com)
2 points by mgreg 3y ago|0 comments
22.
by mgreg 3y ago
Very cool and thanks for sharing. To me a killer feature would be easily running different models simultaneously such as one for embeddings and another for completion (e.g. Chat). This likely can be done already by specifying the model par
23.
by mgreg 3y ago
I'm actually using Ollama for it's Rest API endpoint. Llama.cpp does now have it's server implementation. Unfortunately they do have different endpoints and behave a little differently. * https://github.com/
24.
Justice Department and Google spar over public access to antitrust trial files(theverge.com)
3 points by mgreg 3y ago|0 comments
25.
The first week of US v. Google – Defaults are everything and nobody likes Bing(theverge.com)
214 points by mgreg 3y ago|186 comments
26.
by mgreg 3y ago
Thanks to the folks at MLCommons we have some benchmarks and data to evaluate and track inference performance published today. Includes results from GPUs, TPUs, and CPUs as well as some power measurements across several ML use cases includ
27.
Google’s cookie-replacing Privacy Sandbox reaches major milestone(theverge.com)
3 points by mgreg 3y ago|0 comments
28.
by mgreg 3y ago
That is a very good idea. Any thoughts as to why this hasn't happened already?
29.
by mgreg 3y ago
This is a very good point. There are many other ways to use OpenAI GPT including Bing Chat, Perplexity, Phind, You.com, to name just a few. Some of these services focus on the coding use case as well and may be more fit for purpose for th
30.
by mgreg 3y ago
That's a good point. A parallel can be drawn with AWS / Amazon perhaps. After all FB uses a lot of AI in their products already so you could argue they probably know how to run them in production at scale. Following the Amazon

More ›