Ross Fledderjohn

I got tired of copying content out of Claude, AirDropping it to my phone and uploading it to Instagram just to post once. So I built Wavos.ai. It connects your Instagram to Claude or ChatGPT through an MCP connector. You make the post in the chat, tell Claude when to schedule it, and it goes out. It also sets up comment and DM automations, so when someone comments a keyword they get a DM. It's $5 a month right now for launch. What feature should I add next? Leave it in the comments.

Your CTA is costing you clicks and you do not even see it. Buy now. Learn more. Sign up. These buttons say nothing, so visitors scroll right past them. The fix is not a redesign. It is a rewrite: – Buy now → Get yours today – Learn more → See how it works – Book a call → Find a time that works – Sign up → Start your free trial – Try it → See it in action All 10 swaps are in the slides. I already changed the hero button on my own landing page this month. Same offer, clearer click. Which button on your site is the vaguest right now? Follow @rossfledderjohn for more AI related news

Anthropic just filed to go public, and the most interesting line in the S-1 isn't a revenue number. It's an admission: the world "has only begun to use the capabilities of today's frontier models." Over half the Fortune 500 now pays Anthropic $100K+ a year. On average, that's still less than 0.1% of their total expenses. Users who don't touch Claude Code spend under $75 a month. The reason isn't the model. Anthropic names it in the filing: retraining, process redesign, new policies and culture shifts. It's committing $100M to a Claude Frontier Academy just to train engineers who can get AI into production. The bottleneck has moved from the tech to the people running it. That's where the opportunity is now. Follow @rossfledderjohn for more AI related news

88% of VC-backed founders walked away with zero. Not 88% of the ones who exited. 88% of everyone who raised. Economists at Wharton and the University of Zurich (Azevedo, Scheuer, Smetters and Yang) went through 157,669 US startups and 248,454 founders from 1987 to 2025. In 81% of cases the founders get wiped out as a group. Even when there's an IPO or acquisition, any one founder has a 39% chance of having left before the payday. Meanwhile the top 1% take 79% of all the value. The company's exit and your exit are two different numbers. Know your liquidation preferences and vesting before you sign. Would you still raise knowing these odds? Follow @rossfledderjohn for more AI related news

OpenAI just cut the price of a model that can use your computer in half. GPT-6.1 Terra is $1 per million input tokens and $5 per million output, half of GPT-6.1 Sol. It reads screens, clicks, types and works across apps, and it's live in ChatGPT Work, Codex and the API. Early tests are wild. One builder had it fill out a long web form in 53.5 seconds while Sol took over 3 minutes. But cheaper tokens don't always mean cheaper tasks. On Artificial Analysis, Terra used nearly twice the output tokens Sol did. And OpenAI's own advice is to use Sol or Astra for complex coding and deep research. Would you switch your agents to Terra, or stick with the bigger models? Follow @rossfledderjohn for more AI related news

The best AI prompts never get shared. Fompter by wajo.ai fixes that. Post your prompt. If it hits 1,000 uses, you could win $10K. Use code ROSSFLEDDERJOHN for early access. Comment “FOMPTER” and I’ll send you my Reddit prompt and the link to sign up