Start something only you could build.
London, England
Joined April 2019
- Tweets1.8K
- Following546
- Followers851
- Likes7.1K
Pinned Tweet
A lot of people lately are independently coming to some version of the conclusion ‘the most important thing for humans to do in the AI age is to figure out what is worth building’.
The second order effect of this is a renaissance of psychology/philosophy/literature/art/music as people spend more of their hours returning to the age old questions of the human condition and what constitutes a good life.
You need to discover good values, virtues and goals in order for your agents to help you reach a life of flourishing.
I feel the same is true for entrepreneurship. 1-1 guidance will do wonders for new entrepreneurs.
I truly believe conversational AI will drive the biggest educational leap of our generation. 1-on-1 tutoring is the gold standard for education, but it doesn't scale. AI can change that with truly active conversational learning.
Annie and team set out to solve this with @gakkuai, and they've done a great job. Gakku is an AI mentoring app for K-12 students. Rather than doing the work for them, the voice chat coach guides students through 10 min essay writing "missions" that cover brainstorming, drafting, and revising. It actively trains kids to generate their own ideas, structure logical arguments, and articulate themselves clearly.
If you have school-aged kids, I highly recommend checking it out: gakku.ai
Forced entrepreneurship is coming.
Feeling a massive recession taking hold in the IT freelancing world. It’s been downhill since June.
Old contacts reaching out asking if I have spare work. I’ve started doing the same.
Much less work from existing clients as they save costs because of their own struggles (and try to use AI to do things free/cheaply)
Meanwhile me and my wife have cut back on everything. Going out, eating out, groceries, new stuff for the house. Just trying to hold on to savings.
Anyone else?
In a world without agents, the average person does not need a personal AI agent.
In a world with persistent agents everywhere, everyone will want a personal AI agent.
Kind of like the telephone. Once agents are ubiquitous we’ll see all the obvious ways they’re indispensable.
"Most people aren't looking to save time, they're looking for ways to spend their time."
9 of 15 consumer internet categories have zero AI products in the Top 100. These built some of the biggest companies of the last two eras:
- Streaming
- Social
- Dating
- Gaming
- Travel
- Retail
- Finance
- Real estate
- Jobs
More charts in our Top 100 Consumer AI Apps breakdown: a16z.news/p/top-100-consumer…
I want to rebuild everything I built last year with new models, and by this time next year, I'll want to rebuild it all again.
Clearly this means I need a new strategy.
Here's what I've come up with so far:
1. Don't build to solve problems. Build to fulfill a purpose that will still be a core human pursuit/experience in 10 years. Assume that all of today's problems within that core experience will be solved (and replaced by new ones).
2. Picture the most ambitious, 100/10 version of your idea. Something that would be impossible to build (almost laughable) today. Plan everything around that version, starting today.
3. There's no such thing as a product that will last 5 years. Not even 6 months. You are not building a product to last, you are delivering a frontier experience with the best available capabilities for your stated purpose. It must be capable of refreshing constantly; avoid any complexity that interferes with this.
4. Rebuild your entire product from scratch with the latest models every 6 months. Resist sunk cost fallacy, the token spend for the same intelligence will reduce rapidly. You don't have to actually move active customers over to that version, but you do need to rebuild it and decide once it's built.
5. Reassess your strengths relative to the new AI-assisted world. Some strengths will no longer be significant advantages. Others might now be accessible for the first time (due to other weaknesses being covered by AI assistance). Update what you know about your strengths in the world we're actually living in (markets are relative, so find your comparative advantage).
We'll check back in 12 months to see how this all pans out.
Watching my 7 month old son fixate on a different movement each month (roll, crawl, stand) it becomes very obvious that human development is front-loaded with (mostly?) physical intelligence.
Not sure what to make of this in the era of LLMs.
X has never rewarded low effort, unoriginal projects, because it is primarily about the frontier.
AI hasn’t changed that. We just took a minute to recalibrate, and so people are now realizing that low effort outputs look very different today than they did 2 years ago.
High effort, frontier work is still what we’re all here for and AI will enable a lot more of that, too.
Same principles apply to how you design your work life, btw.
The game design term "chocolate-covered broccoli" perfectly describes most edtech.
You take something boring, wrap it in something fun, and kids see through it in about five minutes.
Meanwhile Factorio teaches supply chain optimization and never has to disguise the learning. The learning IS the game mechanic.
If your educational game needs a fun wrapper, the game design is the problem.
Oh no I'm doing that thing where I write an essay, extrapolate to the limit and lose motivation to take action today because the end state (unreachable for me today) will not require the intermediate state (what I can work on today) at all.
Humans do not compete with AI in the true sense until those systems have the memory, tools and budget to persist and form/pursue their own goals independently. Until then, we are still just competing with other humans.
