Tim | ONLYUP™ 💹
15.7K posts

Tim | ONLYUP™ 💹
@timonlyup
I sometimes have good ideas

People romanticize all-nighters too much. Sleep deprivation is not academic excellence.









Here's a grade 3 explanation for the slow ones Why AI sucks 1. AI is being shoved down your throats every single day. Huge marketing budgets from the biggest company's in the world who are literally force feeding AI to their own workforce and seeing worse results/more mishaps. Ask urself why these company's are losing so much money while simultaneously spending so much money on trying to get people to actually use it. CEOs who don't understand shit and are not in the trenches see an AI post on X, start to feel like they're falling behind, and bark at their employees to start implementing AI - not realising that in most cases it just creates extra work for the employees with more revisions and decreased productivity and more costs 2. AI hallucinates and it can't be fixed. Because LLMs are probabilistic, they hallucinate. So if an 'agent' has a 95% success rate on a single task - which sounds great - chaining 10 tasks together drops the overall success rate to about 60%. In work settings, that compounding failure rate makes true autonomy impossible. The AI company's (and normies alike) defend this by talking about something called 'scaling laws' This is the idea that making the models bigger automatically makes them smarter. But we're running out of high-quality human text to train on. And recent updates to models are demonstrably worse than older models. For copywriters and marketers, this means AI is a great brainstorming and drafting tool. But it can't replace the strategic thinking, empathy, and reliability of a human expert. 3. Zero Corporate ROI Even the corporations pushing this technology are realising it isn't the magic they were promised. A PwC survey of over 4,500 CEOs found that 56% of them have seen zero financial returns from implementing AI. And the rush to replace human workers with AI is already backfiring. Gartner predicts that by 2027, half the companies that fired workers to replace them with AI are going to have to hire those workers back. There's also massive quality control issues when AI is given too much autonomy. In April 2025, Microsoft announced that 30% of their code was being written by AI. Since that announcement, they've experienced more software failures than ever (like a windows 11 update that bricked devices). They literally appointed a new "Engineering Quality Head" to clean up the mess. 4. The business model isn't profitable These companies are literally subsidising the cost of AI to acquire normie users like yourself. They're running at a huge loss to get you hooked on cheap tools. But eventually they will have to raise prices insanely to survive. It's insanely expensive and uses up a ton of energy to power these LLMs. 5. The idea of 'Agentic AI' is a transhumanist psyop. Only god can create true consciousness/intelligence. You can't hold that belief while also believing that these heavily subsidized company's that are circling money between each other to stay alive are going to create an intelligent form. It's a probability based text generator that gets fed material from the internet and predicts the best answer to give you. It makes shit up. It hallucinates. People have been claiming to have cracked human level intelligence since computers first came out. Yann LeCun - the Chief AI Scientist at Meta and the guy who invented convolutional neural networks, which is the foundation of modern AI said publicly that the current architecture is reaching its peak. In his words its a dead end for achieving human-level intelligence. Final thoughts The markets are simply not reflecting reality right now. It's a weird time to be alive And for what its worth I am not 'Anti AI' or against using it. In fact - as some users pointed out on my original tweet - I have been using AI since 2022. I was playing around with it within a couple months of chat GPT release. By mid 2023 we were using it to deliver projects to clients at my agency. I was the first to sell an AI copywriting course (i think). We taught people how to prompt AI and copy chief it to deliver to the same level as a human copywriter in less time. OG followers/newsletter readers remember this. Yes this was a big leap in my business. We could write copy with more speed. I literally held off on a copywriter hire because my senior CW told me he could handle 3 more accounts as a result of AI. As of today I use AI within my business for the following: - Data analysis across 62+ accounts (we have strict QC systems to avoid hallucinations) - Forecasting revenue/profit across businesses - Modelling out my investments and networth - Organizing my thoughts - Writing copy - Market research for new copywriting projects (still, my biggest light bulb moments are from manual research that I feed to an LLM) - Scheduling campaigns/build automations for eCom clients (need to assess if its actually improved efficiency though) - it can take a designed asset and upload it to Klaviyo in HTML form and then schedule it to a corresponding segment with the right subject line and date - My designers use it a lot for image generation. But all those people still work for me lol. The difference is that I was writing 9,000 word VSLs before the average AI user was even introduced to the online marketing space... So me writing copy with AI vs you writing copy with AI will produce a completely different outcome. Plus - I still have 30 people working full time for me. They use AI to help them be more efficient. But their foundational skillset still dictates the quality they output. The ones with more attention to detail and more care produce better work, AI or not. The A players deliver before deadline and the B players deliver last minute. That will always be the case And that's not all The students in my private copywriting community use AI to produce: - Market research documents - Big ideas - Unique mechanisms - Entire funnels - Advertorials/VSLs/Sales letters - AD scripts - Long form statics - In-feed VSLs I actually think copywriting was the industry that benefitted the most from AI. And video/image generation too. But the demand for it hasn't reduced. We've gone from 28 clients -> 60+ in the last few months and its only increasing. If I don't cap our roster (I will) we'll have 100 clients within the next 90-120 days. The students in my community are closing clients every single day. If you're in eCom, you're probably desperate for creative strategists/cold traffic copywriters right now. Every single day there's new job posts from online business owners who are looking to hire copywriters and creative strategists. The notion that AI 'replaced' copywriters literally hasn't happened. Even though you've been told every year for the past 4 years that its coming. All it did was lower barrier to entry and lower the average in terms of skill level. My point is that AI is useful and definitely worth investing mindshare in. But most people are delusional about it and just want to seem like they're ahead of the curve by posting about it. At the end of the day the customer/client cares about the output, not the system you used to produce the output. I would be worried if I dropped everything to go all in on an AI slop offer like selling AI systems to businesses vs putting time into a real valuable career/skill











