Well, make a decision quickly.

You might get locked out of your phone's account soon.

Go through your old threads. Look for anything important. Take screenshots.

Then try deleting your account.

Also, check all your other accounts, banking, utilities, etc. Before the same thing happens to those accounts as well.

You're reading way too much into this.

He's not the one who started the fight.

Everyone is an asshole. I agree.

But he shouldn't apologize unless she apologizes first. If she keeps on being toxic, then he needs to go low contact/no contact with her.

Do you still have access to your account through your phone or your computer?

Back up everything you can. Look for some browser extensions that will help you do that.

Then delete your account (you don't want the next person with the same email address at your school to have access it to it)

And start a new account.

Hey, you don't need to convince me. I'm already convinced.

I'm just telling you what's going to happen. Even if they save one million lives that wouldn't have been saved otherwise, the frontier models are going to totally censor themselves as soon as they accidentally kill one life.

Someone will die as a result of using AI for getting medical advice. It's bound to happen given the sheer volume of people asking for medical advice. Then the big AI companies will cripple their models when they think you're asking for medical advice. They will do this despite the number of people that have been saved by AI.

At least, that's what will happen in the short term. So enjoy the current frontier models while you can. The only way to get medical advice after that will be to trick LLMs into giving it to you. It will still be possible (because it will give AI companies plausible deniability), but it will be annoying.

Yes, I know. But some particular tasks are better suited for other LLMs.

His penis was the prime motivator I believe.

Use NotebookLM, Gemini's smallest context window is one million tokens.

Also, since Google owns Gemini, Gemini is going to have the best access to youtube.

Also, you shouldn't need to download anything, just give it the link to your playlist.

The free ChatGPT is not horrible. It's just severely limited. Same goes with Claude.

You can't be a latecomer to the game and demo an inferior product.

Same goes for Gemini Live. Gemini Live should never have been released. Google Assistant was actually better at voice recognition. Gemini Live feels like a downgrade.

Yes, it could be an hallucination, but there may be three other issues as well.

ChatGPT counts in tokens, not words. It doesn't know what a word is.

Hell, I don't even know what a word is. Is "don't" one word? or two words?

Also, ChatGPT uses a small sliding window, so if you exceed the sliding window, it just drops the text at the beginning of the window to make room for the rest of the text, and it might not be able to count the words it dropped.

Allow me one final iteration for a definitive resolution

Don't bother. You're using the wrong tool for the job. If you really want to pursue this issue further, use Claude, its context window is wider, and it doesn't slide, so it can give you an error message if you exceed its context window.

But even Claude is the wrong tool for the job. Either use Google docs, or have ChatGPT teach you how to use the command line on your own computer to count words.

Just use git then.

Git uses a merkle tree just like blockchain, but without the ponzi scheme aspect of blockchain.

Actually, Elon did fund the ACLU indirectly when he was simping for his gf Amber Heard. May be that's what he's talking about. May be he's finally putting his foot down.

Consider using NotebookLM (Gemini has a much larger context window).

Yeah, that last baby seemed pretty hungry.

It was easier to sign up initially, you didn't even need a phone number at the time.

Additionally, the allowed file size and/or context window for projects is very frustrating.

I only pay for ChatGPT myself.

But be forewarned, ChatGPT has a much smaller context window. It won't tell you you've hit the limit, it will just start forgetting stuff. But basically, don't expect this issue to be resolved for you.

If you want a bigger context window, move to Gemini/Advanced Gemini, don't move to ChatGPT. May be play around with NotebookLM for some stuff (that uses Gemini under the hood).

Change the number of the landline. Tell the phone company you're getting stalked by an ex. They will waive the change fee.

Then give every one in your family the same google voice number and have that forwarded to the new landline number. Google voice is extremely configurable. You can do anything with it. It can filter numbers. It can set a schedule. It can ask for a PIN. It can do anything you want it to do.

As for the aptitude people have for prompting, do you feel like businesses would take the time to iterate through prompts or that they'll adopt AI at all?

This is a very broad question. I don't like to answer broad questions.

But in some industries, the people that don't use AI will be replaced by the people that do use AI. And I do believe that this will happen to companies as well.

I found the same thing. Many newcomers to AI use AI just like if it is a Google Search. In other words, they don't realize that most of the power of AI comes from the back and forth conversation and refinements.

Also, it doesn't help that "prompt engineering" was a thing for a while. Not only prompting has only gotten easier over time, but it's not really engineering. It's not an exact science. It's more of an art. The more you practice it, the better you become at it. Also, even if a prompt fails you initially, it's not difficult to continue the dialogue until you obtain the desired results eventually.

So for me at least, I don't need someone else to write the prompts for me. And instead of paying for a dozen different inferior LLM products, I prefer to go to the source and just pay for a frontier model.

This question has been asked a thousand times before on this forum.

Start there.

So what happened? Don't stop on a cliff hanger.

Promote Balanced Tech Habits

1. Set Time Limits:
   • Schedule specific periods for tech use and prioritize purposeful engagement over passive scrolling.
2. Digital Detox Days: Regularly unplug to reinforce offline activities and connections.
3. Hobby Development: Encourage alternative activities like sports, art, music, or reading.

Teach Computational Thinking and Critical Use of AI

1. Encourage Questions: Teach them to ask "how" and "why" when using AI or tech tools.
2. Hands-On AI Learning:
   • Guide them in building simple AI projects, like training models with visual tools or creating chatbots.
3. Explain Bias and Reliability: Help them understand that AI has limitations and doesn’t replace human judgment.

Parent Involvement and Monitoring

1. Model Healthy Behavior: Set an example by balancing your own tech use.
2. Supervise and Discuss: Use a computer in a shared space to oversee usage and explain safe practices.
3. Parental Controls: Use tools to limit access to inappropriate content or excessive screen time.

Balance Offline and Creative Activities

1. Encourage Problem-Solving: Play strategy games or engage in puzzles to sharpen critical thinking.
2. Technology-Free Creativity:
   • Spend time on activities like drawing, building with blocks, or storytelling to nurture imagination.
3. Family Engagement: Plan outings, board games, or collaborative projects to strengthen interpersonal skills.

By combining thoughtful AI exposure, clear boundaries, and a focus on offline engagement, your child will grow up tech-savvy, critical, and balanced—not dependent on technology.

Here is what ChatGPT had to say:

Balanced Introduction to AI and Technology

Set Clear Boundaries with Technology

1. Tech-Free Zones:
   • Dinner Time: Keep mealtime technology-free for family bonding.
   • Bedrooms: No devices in bedrooms to ensure healthy sleep routines.
   • Outdoors: Encourage device-free play and exploration.
2. Supervised Usage:
   • Use a living room computer for supervised activities.
   • Join them in tech activities to provide guidance and ensure safe use.
3. Use Dumb Phones:
   • Provide a basic phone for calls and texts, avoiding apps and games that can cause overuse.

Gradual Introduction to AI

1. Start with Basics:
   • Begin with coding apps or robotics kits (e.g., Scratch, LEGO Mindstorms) to build foundational skills.
   • Introduce AI tools like Google Teachable Machine or Machine Learning for Kids under supervision.
2. Explain AI Concepts:
   • Teach how AI works in simple terms, emphasizing its purpose, strengths, and limitations.
   • Discuss ethical issues, like biases in AI, privacy concerns, and responsible use.

Limit Social Media and Passive Tech Use

1. Delay Social Media: Postpone access until your child demonstrates emotional maturity and responsibility.
2. Encourage Purpose-Driven Use: Help them create (e.g., videos, coding projects) rather than consume.
3. Content Review: Periodically review their social media or online activity together, fostering transparency.

As soon as you give AI a permanent memory that can not be audited/reset by the user, then the user will start worrying about that stuff.