Sorry for the laggy video camera 🥲. I am currently visiting family while still doing 2 videos/week and the mobile camera failed me here. 95 % of this video is about the screen recording though (which is flawless) so I figured it would be worth sharing anyway! Hope you find a Claude use case here that helps you out ❤.
I know that Chatgpt cannot create a working python program unless I got some basic knowledge to read the codes. Are you saying that Claude 3.5 can literally fix it from scratch?
@@ekot0419 Yes I would say that this jump is so big that it's worth a try. Use the projects feature to store your code though and you might just be surprised. No guarantees though
@aiadvantage I may even fork out subscriptions to try it out. I got stuck for weeks trying to get Chatgpt to create python code from scratch. I can dream about chatgpt apologizing to me all day long. Thank you so much for sharing. Appreciate you.
@@aiadvantage By the way. I am really worried about how fast it advances. The day is coming. I just hope that I can make my bucket of gold before it's too late. Cheers.
API services from OpenAI (for GPT) and Anthropic (for Claude) allows for virtually unlimited interactions, billed based on usage rather than a fixed number of prompts, which can be more cost-effective for heavy users compared to subscription plans as they only pay for what they use, and their access is not restricted by a set number of interactions like plans with prompt limits.
I made a fully functional typing test. I'm not a developer, but it's amazing that I can have an idea and, with a bit of patience, create a piece of software that I can use for myself!
I have the pluss of both and... chatgpt is what I turn to most but having the really long discussions and projects is sick. That bouce engine for chatgpt is stupidly real
Hi Igor. Arnab here . Today’s post was awesome. Though I got too enthusiastic and maxed out on context from Claude with a huge pdf to convert into an interactive website, I’ll get there . The use cases and the notion site is invaluable . Thanks and have a great weekend . Let me know if you are in Amsterdam sometime
I've been using GPT since the beginning and I have made several games with it, but I will admit the process was long. So I recently decided to give Claude a try just fir the coding aspect of it, and I have to say once I got the hang of it I can tell you that when it comes to coding it gets to the finished product faster than GPT with more concise code. I still have to take some codes from Claude back over to GPT simply because GPT can right them out in a longer form. I've also taken some old GPT code that I could not get to work to Claude and it fixed it with ease. If Claude could add some of the attributes of GPT like longer text output, images, etc... It would be the winner. GPT is still tops for me, but I pay for both and I use them both but I get more overall from GPT. Great video...
Thanks for sharing! I also have both now but for code related tasks I find myself going to Claude first every time. The projects feature works really well once you start building something larger too. Just add the various code snippets as uploads and keep on prompting.
@aiadvantage That's what I've been doing, but GPT has me spoiled with just being able to write the full code by just hitting the continue button. I've been using Claude for about a week and I'm getting the hang of it. Thanks again...
when I am asking claude sonnet or opus to create the cat image I am getting an apology from claude: "I apologize, but I'm not able to create, generate, edit or produce images of any kind, including 8-bit style images. However, I can describe what an 8-bit style image of a cat with a hat might look like, or I could try to create a simple ASCII art representation using text characters. Would either of those options interest you? If you'd like an ASCII art representation, I could attempt to create that for you. Or if you prefer, I could describe in detail how an 8-bit style image of a cat with a hat might appear, which could potentially help you or someone else create such an image. Let me know which option you'd prefer, if any." I am even using the paid version...
1st usecase: to show data in a dashboard there is actually pretty solid software on the market. No need to code any shit by your own 2nd usecase: I hate story-telling, it just fuzzying the facts you search for 3rd usecase: yes, PDFs should be forbidden for transport of statistical data, you'd have to check each number manually anyways 4th usecase: nope. But fun-fact, that there was some dude, writing a summarize algorithm like "take not the first, but the second sentence of the text" and sold it to yahoo for millions :/ But yes, good entry point for starting a project maybe, when giving it multiple tries and do the homework.
Am I the only one who CANNOT get claude to call files from another artifact to use as an asset in a script (like you did with the cat SVG in the game artifact)? It just simply doesn't work for me, and I can't have dependencies between two scripts in the same project/conversation. This has felt like the biggest limitation to sonnet 3.5 (yet also seemed like it'd be trivially easy to fix, just make it treat all the files in a conversation like part of a real coding project and allow them to interconnect in different ways) but now I'm seeing you actually being able to call a previously-made SVG into a different script successfully... how did you do that?
And, “make it better” is going to make some random changes. Maybe okay if you have no idea what you want, but using a prompt that explains exactly what should be better, is going to produce better results.
😂 good comment. So true. Saying that to a LLM is very trusting and optimistic? Let’s just spit at the wall and see if it sticks 😂. Seriously though. The LLM has a lot of knowledge so it probably does “make it better” … as long as it’s having a good day. I wonder if Sonnet is as moody as gpt? 🤔. GIGO?
@@arpitshobhasaria1378 yes of course for instant i need to do literature review with other gpts such as scholarly and others i can easily find a paper or related text in the document i have . It can Analyst the data I produce with spss and tell me the results. it can scan my essay give me feedback or draw charts Gannt chart etc cloud does not have any access to any paper thats my use case . All of these might not be accurate but it helps and give you a starting point.
Don’t believe the hype…it’s special because it’s on every session not what it does….we have been generating functional code in home labs quite well…the drama was multi model hosting slowed us down but right now we have been coding and self debugging for a while…the problem is we can’t make a product only sell services because our code is all stolen or may be we don’t know the sources
@@Innesbman I was playing fantasy rpg and when I try to fight or harm anything isn't monster just says it can't continue... man I wish there was a less filtered one but claude 3.5 sonnet us better than most other free ones I used
I will actually donate to your channel if you could create a video on website creation utilizing wordpress and seo best practices to actually make it a valuable website creation tool that will actually show up in the 1st page of google organic results.
Is it too late to cease Ai? Remember Hansel and Gretel? Being tricked? With irresistible goodies? Are we to be… laid off by Ai? Ai jobloss for everyone? Swell robotics doing everything? Or worse, imprisoned inside an… Ai new world order?
I understand your concerns, but I think it's important to approach AI with a balanced perspective. While AI will certainly change many industries, it's unlikely to completely replace human workers across the board. Instead, it's more likely to augment and enhance human capabilities in many fields. Remember, AI is a tool created by humans to solve problems and improve efficiency. It's not some autonomous entity with its own agenda. We're still in control of its development and implementation. As for job loss, history shows that technological advancements often create new types of jobs while phasing out others. The key is adaptability and continuous learning. Regarding an "AI new world order," that's quite an extreme scenario. There are ongoing discussions about AI ethics and regulations to ensure its responsible development and use. Instead of fearing AI, we should focus on how to harness its potential to improve our lives while addressing potential challenges. It's not too late to shape the future of AI - in fact, now is the perfect time to engage in these important conversations.
