Tag: openai
-
A process for getting good-enough outputs from OpenAI’s Deep Research
A practical guide to using OpenAI’s Deep Research feature effectively, detailing a four-step process that involves creating prompts with ChatGPT o1, answering clarifying questions, generating comprehensive reports in minutes instead of days, and achieving “good-enough” results that would typically require weeks of research.
-
[Note] Introducing deep research
“Deep research is OpenAI’s next agent that can do work for you independently—you give it a prompt, and ChatGPT will find, analyze, and synthesize hundreds of online sources to create a comprehensive report at the level of a research analyst.” – OpenAI
-
Swarm is a framework for developing multi-agent systems
Swarm is an experimental framework from OpenAI, for building, orchestrating, and deploying multi-agent systems.
-
The significance of OpenAI’s $6.6B investment round
OpenAI’s unprecedented $6.6 billion investment raise suggests they demonstrated something remarkable to investors, though not necessarily just raw intelligence gains. Whether it’s improved safety, efficiency, or multimodal capabilities, this massive vote of confidence hints at breakthroughs we haven’t yet seen publicly—developments that could reshape AI’s integration into society.
-
Being inaccurate isn’t the same as being useless
New research on AI model factual accuracy shows that while language models struggle with certain difficult questions, this doesn’t diminish their value as thinking partners. Like human conversations, where perfect accuracy isn’t required for productive discussion, AI’s occasional inaccuracies don’t prevent it from being a useful collaborative tool.
-
Diminishing returns of LLMs doesn’t stop progress
Recent discussions about LLM diminishing returns suggest OpenAI’s next frontier model may not be significantly smarter than GPT-4. However, this plateau in intelligence doesn’t diminish the technology’s potential, as improvements can focus on making models cheaper, faster, smaller, and better at specific tasks rather than increasing raw intelligence.
-
Court ruling: Language models don’t copy information; they synthesise it
in AIMasse, B. (2024, November 8). OpenAI’s data scraping wins big as Raw Story’s copyright lawsuit dismissed by NY court. VentureBeat. The judge noted that “the likelihood that ChatGPT would output plagiarized content from one of Plaintiffs’ articles seems remote.” This reflects a key difficulty in these types of cases: generative AI is designed to synthesize…
-
Frameworks GPT helps you think through problems
Ethan Mollick’s Frameworks GPT helps you work through difficult problems by suggesting suitable frameworks to help structure your thinking.
-
Good (non-technical) overview of how LLMs are getting smarter
Ethan Mollick’s non-technical overview of the two scaling laws describing how generative AI models keep getting smarter.
-
Learning to Reason with LLMs
OpenAI o1 is much better at reasoning through problems.
-
OpenAI releases their ‘Strawberry’ language model
OpenAI releases OpenAI o1, codenamed ‘Strawberry’, in response to common reasoning problems inherent in language models.
-
More to AI detection than accuracy
AI text detectors, like OpenAI’s 99.9% accurate tool, aren’t the solution to academic cheating. These detectors have limitations, including model-specific detection and manipulable statistical features. We’re not going to find answers by entering into an arms race with students, by trying to build increasingly accurate AI detectors.
-
AI tutors are getting very good
I know the idea of AI replicating some parts of the function of a tutor isn’t that comfortable, and there are whole rafts of the more human aspects, such as emotional intelligence, that this work doesn’t go near. But also, we know many students value AI for learning. They value its availability, patience, and lack…
-
What OpenAI did
https://www.oneusefulthing.org/p/what-openai-did With universal free access, the educational value of AI skyrockets (and that doesn’t count voice and vision, which I will discuss shortly). On the other hand, the Homework Apocalypse will reach its final stages. GPT-4 can do almost all the homework on Earth. And it writes much better than GPT-3.5, with a lot more…
-
OpenAI makes GPT-4 available to everyone for free
https://openai.com/index/spring-update This is a big update that (probably) pushes ChatGPT back into the clear leader spot. Omni’s voice, audio and video capabilities are interesting, but I think the bigger news is that OpenAI is now making GPT-4 level generative AI available to everyone for free. Considering that the free version of ChatGPT was using GPT-3.5,…
-
BIP AI – AI fundamentals and prompting
In this workshop for the Blended Intensive Programme on AI in education and research, Antonio Lopes, Hugo Santos and I introduce generative AI models and techniques for effective prompting. Participants explored responses across chatbot platforms to compare nuances in AI outputs. The workshop provided a practical foundation for understanding and harnessing generative AI capabilities.
-
Stop using AI detection services because they don’t work
The researchers conclude that the available detection tools are neither accurate nor reliable and have a main bias towards classifying the output as human-written rather than detecting AI-generated text. Furthermore, content obfuscation techniques significantly worsen the performance of tools. Weber-Wulff, et al. (2023). Testing of detection tools for AI-generated text. International Journal for Educational Integrity,…
-
My thoughts on the different generative AI tools I’m using
TL;DR Here are is my ranked list of suggestions, based on my own experiences and use-cases: Over the past year or so, I’ve been experimenting with a few different language models and image generators. Over time, I narrowed in on Claude, and wrote about my preference for using it over other options. A couple of…