It’s great to see the FTC is looking into this. Every time I’ve had to deal with Adobe subscriptions it always feels like they put you through some maze which usually ends in being forced to pay a cancellation fee or losing in some other way. I recently lost 160 asset tokens I had saved up in Adobe Stock after canceling my subscription. I’m sure it mentions this in the fine print somewhere but the only reason I had saved that many tokens up was because I couldn’t cancel without paying the fee. So they trapped me, took my money and then took back the tokens I only paid for because they locked me in with a cancellation fee in the first place. At the very least I should be able to keep the asset tokens. The whole thing seemed absolutely ridiculous to me.
There is only a cancellation fee if you buy an annual subscription, paid monthly, and then you don't pay for the 12 months you committed to. For people who don't want to buy a year at a time there is a monthly plan that is more expensive since you are no longer buying in bulk.
It’s not about prompt-fu (granted there are some very powerful prompt-fu things you can do if you’re feeling creative :)). I think it’s more about the way you’re framing it’s use case. It will not do the work for you reliably, but it absolutely will allow you to learn what you need to do the work at a rate 100x faster than you could otherwise.
I find it’s very good at correcting your own mental models. For example, if you’re unsure about you’re mental model of how a piece of technology works, explain your mental model to the system and ask it to check your understanding. It’s amazing for using it as a real-time sounding board for checking your understanding of how you think things work.
that's the catch though if it's something I already know then I can already stop when It's bs me but I also don't really need it then.
when I don't know about something then how do I now it's bs me ? double check if work, which mean doing the work the work twice once with the LLM and another with web search.
the problem is that as I could tell from my trials is that these LLMs are very good bullshiters so it make it hard to discern in a fast fashion.
one thing I did find it useful for is asking vague questions and then using the keywords in the response to start my web searchs with.
I wonder if the people who haven’t discovered it’s usefulness yet haven’t applied enough creativity into their own inputs. The output can definitely be underwhelming when you pass a cookie cutter input and expect an earth shattering output. The power of the system is seemingly in the fact that your input options are literally infinite. LLM’s are really not dissimilar from computers in that sense. You can write your first program that adds 5+5 and prints 10 and walk away from that experience thinking “Computers are overhyped. I don’t see what the big deal is.” But then you can also build Figma, Call of Duty, and cancer prediction algorithms.
I’ve found these models to be the single most valuable thing I’ve ever come into contact with in my life. My ability to learn complex subjects is a world apart from what it was before ChatGPT. I can study a topic and then have hours of conversation with the model through the voice based interface on the iOS app. The ability to ask any question about any topic in your own words and instantaneously get a response from a machine that’s likely been trained on every text book ever written is absolutely insane. Any time I’m reading a paper or documentation and I don’t quite understand something I can ask the question in my own words and get an immediate response. You can ask the model to explain it visually, analogize the concept to a concept you’re already well versed in, have it ask you questions to challenge your own understanding, literally anything.
There’s certainly a hype train around LLM’s and many specific aspects of what’s claimed is completely over stated, but if you’re a knowledge seeking human and you’re not using these models as a tool to learn then you are as an absolute fact asleep at the wheel.
> I can study a topic and then have hours of conversation with the model through the voice based interface on the iOS app. The ability to ask any question about any topic in your own words and instantaneously get a response from a machine that’s likely been trained on every text book ever written is absolutely insane.
It's also likely trained on nearly every website ever, including the wildest of wild conspiracy theories and bunk science. Try asking it detailed questions about something you would consider yourself an expert in and see how often it gets details completely wrong.
One nice way of handling this is to use the CSSStyleSheet object. In a nutshell, you can create a module like shadowCssRules.js and in that create a CSSStyleSheet object and set your global rules ( * { box-sizing: border-box;} ) then in your WebComponent constructor you call this.shadowRoot.adoptedStyleSheets = [shadowCssSheet]. This MDN page has an example at the bottom https://developer.mozilla.org/en-US/docs/Web/API/CSSStyleShe...
Also allows you to update the style sheet from anywhere in your js and it’ll effect all the shadow doms that adopted it
Adobe Illustrator and other infinite canvas platforms provide a compelling case against this. Forcing yourself to draw pictures of what you’re taking notes on especially in studying software systems where the ideas are abstract and not necessarily visual is really useful. The added ability to nest concepts inside other concepts like Russian doll note taking is also pretty interesting. You can also screen clip whole articles and paste them next to your drawings and notes. Infinite canvas note taking is probably the most useful thing I’ve ever discovered in terms of learning technical systems.
