🗞 What's New: Everything you need to know about OpenAI's new model

This week, OpenAI released its new model, o1:

• These models can reason about complex tasks, and solve trickier problems than earlier models. Play around with it to discover what to build.
• The true cost of Shiny Platform Syndrome is spreading yourself too thin. Exercise restraint, and focus on the great thing you're already doing.
• Dodging hurricanes and launching MVPs. Thomas Wink saves money by sailing full-time and working from his boat, and just launched MARQT AI.

Join Intercom’s Startup Program and get 12 months free! Create a direct line to your customers with the only complete AI-first customer service solution. Join thousands of other fast-growing startups using Intercom, including Coda, Qonto, and Pitch. #ad

by Stephen Flanders

This week, OpenAI released its new model, o1. Here's everything you need to know about it!

What is o1?

OpenAI o1, previously codenamed "Strawberry," is a new series of AI models that can reason through complex multi-step tasks.

They're designed to spend time thinking before responding to your requests. They can reason about complex tasks and solve trickier problems than earlier models, especially when it comes to science, coding, and math.

This means o1 is much better at translating a corrupted sentence, coding a video game from a prompt, and perhaps most impressively, answering how many "R's" are in "strawberry."

How does it work?

o1 uses a mix of chain-of-thought and reinforcement learning to reason through problems. At a high level, this just means that it breaks down problems into simple steps, then goes through those steps until it reaches the right answer. If it hits a snag, it simply tries a different approach.

Unfortunately, OpenAI hasn't published its research, but those in the know are hypothesizing that there's a separate critic model scoring outputs with a heuristic.

Why does it matter?

Up until now, reasoning was impossible for AI models. They were mere "stochastic parrots," trained on immense amounts of data that enabled them to predict the correct answer to simple questions. Now, they're capable of thinking through questions in much the same way that a human does.

That's a whole new level unlocked, which OpenAI itself acknowledges (hence the name o1, instead of GPT-5). In the short term, this likely means that true AI agents are possible. In the midterm, it may be able to invent new cancer drugs or breakthrough batteries. In the long term, OpenAI may have found the pathway to AGI.

What should indie hackers do?

First, play around with it. Odds are that it can help you with something. So, get in there and start using it.

By doing so, you'll learn its strengths and weaknesses, and thus, what you can build with it. Countless startups were born from GPT-4, and this is the biggest AI advance since then.

Now is not the time to stay on the sidelines!

Discuss this story.

This issue is sponsored by Agent.ai

Join a community of professionals who are outsmarting the competition by leveraging the power of AI to GSD.

Hire AI agents to automate tedious tasks, gain deep insights, and drive better results. Free up your time to focus on what matters most.

Get started with 100 free credits!

by Jay Clouse

Every now and then, you may feel the pull to start investing time in a platform that you're not currently active on. Maybe it's YouTube, Bluesky, or X. It could be anything, but whatever it is, it's probably a trap.

When we have limited experience with something, we tend to overestimate our ability to do that thing. Tim Urban of Wait But Why created this visual:

When you're not yet publishing on X, Bluesky, or wherever, your feet are planted firmly at the top of Child's Hill. Since you don't have much experience with that platform outside of being a consumer, it feels easy. In fact, it feels easier than the platforms you're currently tackling.

The vicious cycle of Shiny Platform Syndrome

We fall prey to Shiny Platform Syndrome because it's exciting to start something new. Something new has seemingly limitless potential, and because we don't have any data yet, our imaginations run wild with what could possibly happen for us.

To further complicate things, we tend to get a (false) positive response when we launch new things.

The real cost of Shiny Platform Syndrome is that, by spreading yourself thin across more platforms, it's almost guaranteed that your work will suffer on each of them.

When you fall prey to Shiny Object Syndrome, you're generally signing up to become:

1. Bad at something new,
2. More mediocre at the things you've been practicing over time, and
3. Burdened by a new commitment.

Don't fall into the Shiny Platform Syndrome trap. If you want to create something that people truly love, exercise restraint, and focus on that thing!

This article was originally published here.

Discuss this story.

from the Trendy Software Ideas newsletter

👾 Video game companies have been hit with an EU complaint over "tricking consumers."

💲 Listen to the world's knowledge on Overlap (YC S24). #ad

🔎 Non-obvious signs of early traction, and how to spot them.

💻 More Millennials and Gen Z say they feel more like themselves online than offline.

📈 How to improve retention.

🦾 The bot keeping humans out of dangerous silos.

Check out Trendy Software Ideas to discover more software ideas inspired by recent news.

by Anugraha Sundaravelu

For the past three years, Thomas Wink has been living the digital nomad dream, sailing and working from his boat.

Just weeks ago, he finished the MVP for MARQT AI, while crossing the Atlantic Ocean on a nonstop solo trip from Colombia to the Azores.

It all started with boats

Thomas started his indie hacking journey in the early 2000s, when he set up his online chandlery. He began selling nautical items for ships and boats. Using his dad's connections in the boating industry, he started getting supplier contracts and filling the catalog with products. Although SEO was still quite new then, within two weeks of going live, he received his first orders from organic search.

After running the business for five years, Thomas acknowledged that he was burning out. So, he sold.

Dodging hurricanes and launching MVPs

MARQT AI started off as a side hustle, and has slowly evolved into a full-time commitment. Thomas teamed up with a good friend who is a marketing specialist, and they also brought on a talented designer.

A few weeks ago, Thomas had to run from Hurricane Beryl in the Caribbean Sea, and almost hit a dead whale while approaching The Azores. It took him four weeks to sail the 3.7K+ nautical mile trip, and by the end of it, he and his team had launched the MVP.

With MARQT AI, Thomas aims to democratize market research by offering a low-cost solution for startups and scaleups, giving businesses a competitive advantage at a time when resources are tight.

MARQT AI offers a subscription-based service, starting with a free trial. As the MVP just recently launched, the immediate focus is on usage over income. The product isn't driving revenue just yet.

Living rent free

Thomas sees sailing as his biggest investment, since he can travel and save on rent.

He also keeps his costs low by looking for bargains, buying secondhand gear, and doing all the boat maintenance himself.

He worked in a shipyard a few years ago to learn more about diesel engine maintenance and repair, which really comes in handy here!

Discuss this story.

by Tweetmaster Flex

I post the tweets indie hackers share the most. Here's today's pick:

Forward it to a friend, and let them know they can subscribe here.

Also, you can submit a section for us to include in a future newsletter.

Special thanks to Jay Avery for editing this issue, to Gabriella Federico for the illustrations, and to Stephen Flanders, Darko, Jay Clouse, and Anugraha Sundaravelu for contributing posts. —Channing