I test ways to make money online. AI, content monetization & side hustles. Every possible way to get money is tested and reported.
Joined June 2026
- Tweets1.8K
- Following348
- Followers341
- Likes1.3K
Pinned Tweet
I created 1 facebook page 1 month ago and for my surprise it has been doing it very well.
only posting reels 1 or 2 times per day.
I hope it can be monetizable soon.
I started earn money training AI with task videos 1 week ago and I can't be more impressed.
I recorded every task I do everyday
-cleaning dishes
-feeding my pets
- water my plants
Getting paid around 6$-8$ per hour living outside of US and 10$-12$ living in US
100% recommended!
This is my affiliate link if someone wants to start earning and also helping me!!
app.claru.ai/signup?ref=3c1p…
I’m starting from $0.
For the next 30 days, I’m turning this account into a public monetization experiment.
The goal:
Take a rebranded X account from $0 to its first online dollar.
I’ll share:
• Every strategy I test
• Time and money invested
• Audience growth
• Revenue generated
• Everything that fails
Keep in touch to don't lose this opportunities!!!
So I started this account 2 months ago and i can't feel more disappointed.
not because of the content and the people but more about the space I placed this account.
Everyone is just creating, copying the same thing over and over again.
thousands of bid websites, people with the hundreds of calorie tracker apps and X growth SaaS.
Those replies with the same format that constantly remind you how these people don't care about being original.
I though that maybe if i shared something that I do everyday (my job) this would be easier because yeah, it's cool to share insights but i becomes nothing when the people around this spaces are just masks.
I love X, but i hate this space.
I think I'm gonna pivot this entire account to something else, feel free to unfollow if you dislike it!
since a few weeks i have been noticing how a bunch of people that were "building something" dropped because the new monetization program is harder.
Now that Claude and ChatGPT uses watermark and Google has started the new anti spam campaign everything makes sense.
Google has been trying to regulate the spam and autopilot sites since 2022.
This means that a lot of people running these sites in autopilot would notice a decrease pretty soon.
After a week full of things and a massive workload.
I'm finally here!!
How is X going? Any changes or something I need to know?
Glad to reconnect with all the people I was talking and share more ideas a conversations!!!
I just went out of X for 1 day, no posting or replying.
20 unfollows and a lot of things happening.
The algo is open source and all that stuff.
Interesting how people is receiving less comments since the new ads revenue started.
Happy to have more people creating value content and stop seeing the farmers in my timeline.
But somehow thid chsnge has encourage people to stop replying.
4250 is a crazy numbers, goddammit.
I would have to spend a whole life here too.
I just wanted X to pay my token credits....
Val | Income Sidequests retweeted
Kimi K3 instead of Fable 5.
DeepSeek Flash and Luna instead of Sonnet 5 and Haiku, while still giving you around 90% of what you’d get from Opus 5, at much higher speeds and a fraction of the cost.
If you want to take a stand against Anthropic’s policies, now is the perfect time.
Cancel your subscription. Use the alternatives.
We can also evade this using open source models, DeepSeek would never do this to any of us.
Another step back made it by Closed models
BREAKING: AI-generated text will be watermarked by Claude.
This is going to be like a "synthid" but for text.
LLMs generate random text, with per-token probabilities.
To select the tokens according to the probabilities, you use a random number generator.
If, for some small part of that randomness, you use the last 10 words as a seed, you can statistically detect this randomness source in any generated text.
Now you can make the effect of the watermarking on the generation arbitrarily small, simultaneously making the watermark easier to remove.
Scott Aaronson came up with this approach while working at OpenAI years ago, but they decided not to implement it. He's written about it on his blog.
New Tool idea: Remove AI watermaks.
This would explode in the following days.
BREAKING: AI-generated text will be watermarked by Claude.
This is going to be like a "synthid" but for text.
LLMs generate random text, with per-token probabilities.
To select the tokens according to the probabilities, you use a random number generator.
If, for some small part of that randomness, you use the last 10 words as a seed, you can statistically detect this randomness source in any generated text.
Now you can make the effect of the watermarking on the generation arbitrarily small, simultaneously making the watermark easier to remove.
Scott Aaronson came up with this approach while working at OpenAI years ago, but they decided not to implement it. He's written about it on his blog.