David Nadeau
@Pythonner
Followers
634
Following
1K
Media
114
Statuses
3K
Full cycle Machine Learning: I design, develop, deploy & support. Bot Populi, Bot Dei.
Ottawa, Canada
Joined April 2009
A recent work from @iddo claimed GPT4 can score 100% on MIT's EECS curriculum with the right prompting. My friends and I were excited to read the analysis behind such a feat, but after digging deeper, what we found left us surprised and disappointed. https://t.co/mpDqlenk04 🧵
flower-nutria-41d.notion.site
What follows is a critical analysis of “Exploring the MIT Mathematics and EECS Curriculum Using Large Language Models”
Exploring the MIT Mathematics and EECS Curriculum Using Large Language Models Presents a comprehensive dataset of 4,550 questions and solutions from all MIT EECS courses required for obtaining a degree https://t.co/cZfjNRu4AL
53
795
3K
This is great. AI warfare presented as good thing is questionable (think lethal autonomous weapons) but else, I agree with all of it!
Why AI Will Save The World By Marc Andreessen The era of Artificial Intelligence is here, and boy are people freaking out. Fortunately, I am here to bring the good news: AI will not destroy the world, and in fact may save it. 🧵
0
1
3
A new Ahead of AI issue is out, where I am covering the latest research highlights concerning LLM tuning and dataset efficiency:
magazine.sebastianraschka.com
In the last couple of months, we have seen a lot of people and companies sharing and open-sourcing various kinds of LLMs and datasets, which is awesome.
8
50
275
Everyone should learn to fine-tune LLMs. You can try (and fail) to force GPT-4 on to every solution, or you can reduce your costs and improve your accuracy by fine-tuning a task-specific LLM. Here's why, and some tips for getting started:
23
91
827
We're launching ten $100,000 grants for building prototypes of a democratic process for steering AI. Our goal is to fund experimentation with methods for gathering nuanced feedback from everyone on how AI should behave. Apply by June 24, 2023:
openai.com
Our nonprofit organization, OpenAI, Inc., is launching a program to award ten $100,000 grants to fund experiments in setting up a democratic process for deciding what rules AI systems should follow,...
531
1K
4K
If you are Canadian and speak French and you have a PhD related to data science *or* networks/security, you can become my colleague. https://t.co/Kw91SjIMsz
1
3
9
Poking an insanely high-dimensional black box with a stick for a living 🪄 https://t.co/JRY0TlqZyF
washingtonpost.com
Proponents of the growing field argue that the early weirdness of AI chatbots can be avoided by a human giving the machine all the right instructions.
3
2
25
2023 is the year AI/ML went from: -> people cherry-picking examples of systems giving the right answers to: ->people cherry-picking examples of systems making mistakes. Tells a lot on the giant leap we just made.
0
1
4
Unashamed bragging: in 2007, on page 69 of my PhD thesis, I explained how to build a kind of language model on a Terabyte-sized dataset. Pre-Deep NN era: https://t.co/JlFur4xPpn
0
0
1
Prompt Engineering Guide (1.8K ⭐️) Organized papers by date so it's easier to track the latest developments in prompt engineering. All the latest tools added too! https://t.co/24k6YQrMcz
27
235
1K
@flybird110013 I have a slightly extended version here, I hope it helps:
sebastianraschka.com
Since the release of the AI Classifier by OpenAI made big waves yesterday, I wanted to share a few details about the different approaches for detecting AI-g...
1
1
5
OpenAI just launched the "AI Text Classifier" to identify texts generated by AI. Tried it, and IT DOES NOT WORK. https://t.co/ibTQboK21D Using my Python ML book published in 2015: 1) @randal_olson's foreword: unclear 2) my preface: possibly AI 3) paragraph from Ch1: likely AI
46
180
933
Open Problems in Applied Deep Learning If you're looking for interesting open problems in DL, this is a good reference. Not sure if intentional but it also looks useful to get a general picture of current trends in deep learning with ~300 references. https://t.co/AXEZQNtCjr
21
228
962
AI community: make sure to review the output of AI, generate alternate images, refine your prompt (else it's often creepy lol). MarketWatch: 'image of a bear in a house of mirrors' noice, publish.
0
0
1
My team and I have availability. Drop me a line at me@afilina.com if you: - Need a PHP or framework upgrade. - Are struggling with old spaghetti code. - Not sure how to introduce automated tests. - Would like to pick my brain or pair program for an hour. Thanks for retweeting.
0
34
33