Liens en vrac de Ramenos (v2)
898 shaares
11 résultats
taggé
chatgpt
ChatGPT toolkit. Quelques bonnes ressources dans cette liste. A explorer plus tard.
Quelques statistiques intéressantes sur les usages de l'IA générative.
Here are some samples from the 100, with one quote for each. The full list is at the bottom of this article.
Generating ideas (#1). “I love it for brainstorming because it’s like the perfect teammate. It can keep up with me and doesn’t get hung up on dead-end ideas, and it can summarize what we come up with so it’s easier to present or reference later on.”
Specific search (#2). “There was a particular cookie my grandmother used to give me and I really liked the taste and texture, and I had looked at the grocery to no avail until one evening … I decided that it might be fruitful to ask ChatGPT for help … It was SnackWell’s.”
Editing text (#4). “I use it to check my own biases with op-eds and speeches and other political stuff. If something makes me feel strongly, I copy it into ChatGPT and ask it to tell me the logical fallacies and possible misinformation in the piece. It is a HUGE gut check!!”
Drafting emails (#11). “I work in investor relations and the amount of time I’ve saved using ChatGPT to help me draft emails is almost unquantifiable.”
Simple explainers (#12). “It’s also way better at explaining concepts to non-engineers than us engineers are. By default it writes at a 5th-grade level, which is perfect for a lot of people we interact with at work.”
Excel formulas (#14). “I have to write a lot of .vb and Excel formulas to reconcile data from less technical people. ChatGPT helps 45-minute tasks take about three to five minutes.”
Making a complaint (#23). “A car wash damaged my wife’s SUV and refused to pay, so GPT drafted a demand letter for me, and I took them to small claims court.”
Generating appraisals (#26). “I know some managers who use it to help punch up performance appraisal write ups for their employees.”
Editing legal doc (#44). “I fed it a long, overly complex service level agreement for a SaaS contract and ask it to rewrite it to make it simpler and easier to digest. It kept the important SLA terms but condensed the language by 70%.”
Sampling data (#85). “It’s great for producing demo data. [If you] need a bunch of fake company names or customer names or product codes, ChatGPT is good at deriving stuff like that.”
Here are some samples from the 100, with one quote for each. The full list is at the bottom of this article.
Generating ideas (#1). “I love it for brainstorming because it’s like the perfect teammate. It can keep up with me and doesn’t get hung up on dead-end ideas, and it can summarize what we come up with so it’s easier to present or reference later on.”
Specific search (#2). “There was a particular cookie my grandmother used to give me and I really liked the taste and texture, and I had looked at the grocery to no avail until one evening … I decided that it might be fruitful to ask ChatGPT for help … It was SnackWell’s.”
Editing text (#4). “I use it to check my own biases with op-eds and speeches and other political stuff. If something makes me feel strongly, I copy it into ChatGPT and ask it to tell me the logical fallacies and possible misinformation in the piece. It is a HUGE gut check!!”
Drafting emails (#11). “I work in investor relations and the amount of time I’ve saved using ChatGPT to help me draft emails is almost unquantifiable.”
Simple explainers (#12). “It’s also way better at explaining concepts to non-engineers than us engineers are. By default it writes at a 5th-grade level, which is perfect for a lot of people we interact with at work.”
Excel formulas (#14). “I have to write a lot of .vb and Excel formulas to reconcile data from less technical people. ChatGPT helps 45-minute tasks take about three to five minutes.”
Making a complaint (#23). “A car wash damaged my wife’s SUV and refused to pay, so GPT drafted a demand letter for me, and I took them to small claims court.”
Generating appraisals (#26). “I know some managers who use it to help punch up performance appraisal write ups for their employees.”
Editing legal doc (#44). “I fed it a long, overly complex service level agreement for a SaaS contract and ask it to rewrite it to make it simpler and easier to digest. It kept the important SLA terms but condensed the language by 70%.”
Sampling data (#85). “It’s great for producing demo data. [If you] need a bunch of fake company names or customer names or product codes, ChatGPT is good at deriving stuff like that.”
Formation introductive à la génération de contenus avec ChatGPT de Vincent Terrasi. Gratuit mais très bien pour démarrer
We administer a Turing Test to AI Chatbots. We examine how Chatbots behave in a suite of classic behavioral games that are designed to elicit characteristics such as trust, fairness, risk-aversion, cooperation, \textit{etc.}, as well as how they respond to a traditional Big-5 psychological survey that measures personality traits.
L'extension ChatGPT pour l'utiliser depuis Chrome ou Chromium
Quelques bon conseils de la part de Vincent Terrasi pour mieux utiliser le prompt de ChatGPT
Des gens s'amusent à converser des heures avec ChatGPT. Ca fait froid dans le dos... Et ça coûte cher en energie !
ChatGPT, l'outil qui permet de mettre en avant les gens qui ne font pas bien leur travail... Un exemple typique dans le domaine du légal. C'est moche et je sens que ça va être le début d'un grand nombre de cas...
Un prompt ChatGPT. A voir s'il est correct ou pas...
Haha je l'adore celui-là. Le bot Feelbetter, pour les gens qui n'ont pas les moyens de se payer un thérapeute mais qui veulent quand même échanger sur leur problème
We demonstrate through human evaluation that existing de-
tectors of machine-generated text are effective at predicting
low quality pages, outperforming, quite surprisingly, super-
vised spam classifiers. To our knowledge, this is the first use
of machine detection for a different NLP task.
• Using half a billion webpages, we conduct the largest appli-
cation of the detection models in the wild.
• We quantify the low quality pages that are surfaced by our
detector models. We perform extensive analysis, breaking
them down by attributes such as document length, age, and
topic.
• We qualitatively characterize and categorize the nature of the
low quality documents. We find traces of essay generation
farms, machine translated text, keyword optimizations, and
Not-Safe-For-Work (NSFW) content.
L'étude est intéressante et permet de mieux comprendre comment GTP fonctionne en arrière.
tectors of machine-generated text are effective at predicting
low quality pages, outperforming, quite surprisingly, super-
vised spam classifiers. To our knowledge, this is the first use
of machine detection for a different NLP task.
• Using half a billion webpages, we conduct the largest appli-
cation of the detection models in the wild.
• We quantify the low quality pages that are surfaced by our
detector models. We perform extensive analysis, breaking
them down by attributes such as document length, age, and
topic.
• We qualitatively characterize and categorize the nature of the
low quality documents. We find traces of essay generation
farms, machine translated text, keyword optimizations, and
Not-Safe-For-Work (NSFW) content.
L'étude est intéressante et permet de mieux comprendre comment GTP fonctionne en arrière.