I can't do everything
In a Nutshell
AI's capabilities have exploded to the point where users face "analysis paralysis" from infinite options, but human bottlenecks—limited time, attention, and energy—remain fixed. The core lesson is "you can do anything, but you can't do everything," requiring deliberate minmaxing: focusing resources on 3-4 high-value projects rather than scattering effort across countless AI-generated outputs. Without active engagement and prioritization, AI simply creates worthless "slop" that accumulates unused.
These notes were generated by AI and may contain inaccuracies.
A tweet posted across LinkedIn and YouTube addressed the problem of choice in the context of AI advancement. The speaker tends to be 3 to 6 to 12 months ahead of others in realizing the lived experience of what's coming down the pipeline. The psychological phenomenon called the problem of choice occurs when more options create mental burden from trying to track the best option, leading to analysis paralysis.
The release of Opus 5.5, Astra, and Soul 6.1 has created an inflection point where models are so capable that "I could do anything" - work on books, turn novels into manga or comic books, animate content, and pursue countless projects. However, limited hours in a day and mental bandwidth require minmaxing around actual desired achievements.
While tasks can be delegated to different AI systems (Grockbot, ChatGPT), the executive function required to review and enjoy results is not fungible. AI can write fanfiction, books, or conduct research, but this only accumulates files unless the user actively engages with the output. Without engagement, it becomes "pointless token spend."
FOMO around unused quotas has been common, but Opus 5.5 represents an inflection point where most users no longer run out of quota. While some maintain multiple accounts (two ChatGPT Promax, five Claude accounts), the bottleneck has shifted from account limits to the user themselves.
The speaker identifies three human bottlenecks:
- Intellectual capacity
- Metabolic capacity
- Clock time and sleep-wake cycles
Even with asynchronous AI processing, verification time and the time to enjoy and use output remain constraints.
AI and automation are creating a "quarter life crisis kind of thing" - not a one-to-one parallel, but similar in nature. The quarter life crisis, typically occurring in the 20s to early 30s (mid-20s for the speaker), involves realizing "I have so much potential. I can do anything" but also recognizing "I can't do everything" and being limited by means such as time, education, money, experience, and relationships.
The fundamental lesson from quarter life crisis is: "you can do anything, but you can't do everything" or "you can be anyone. You can't be everyone." With only one life to live and "one year per year," people must minmax their approach to AI usage.
Minmaxing involves optimizing around specific desired traits, similar to putting zero skill points into charisma and everything into strength when playing an orc character in tabletop RPGs or video games. Applied to AI, this means focusing time, energy, cognitive resources, tokens, and quota on three or four specific things that provide the most value rather than being well-rounded.
The speaker maintains a sprawl of projects including fanfiction, fiction, novels, animations, research, and books. The reality is that "humans have produced slop for a long time" and "AI just turns it into a slop fire hose." The approach requires focusing on what one actually wants to invest significant time and energy into - books that will actually be read or published, games that will actually be enjoyed.
There is merit in experimenting and challenging oneself without requiring every activity to be productive. This aligns with the principle that individuals determine their own values, but requires conscious evaluation of whether activities align with higher-order values and how time is being spent.
People with ADHD have "a billion interests," and AI can indulge every single interest regardless of whether it's a good idea. This may explain why some people experience AI psychosis.
The video was recorded at 1080p and 30 fps settings to optimize bitrate for the phone processor. The manuscript for "Post Labor Economics" is nearly complete following a call with the researcher on Friday, with only proofreading, formatting, and comments review remaining. Voiceover narration warm-ups were practiced before recording.
Keep David Shapiro in your library
Save the videos and channels worth coming back to, and find them again in one place.
