Aakash Gupta Profile picture
Jun 3, 2023 27 tweets 11 min read Read on X
What an insane week in AI 🤯

Unity AI
AI Antibiotic
Minecraft AI
MIT: Math AI
Picture-to-3D
Google Starline
Nvidia Game AI
Google Flood AI
ChatGPT in Court
TIME: Humanity End
Neuralink FDA Approval
42% US Not Used ChatGPT
Japan No Copyright OK for AI

Here what you need to know:
1. Unity AI

Unity, the game engine behind 80%+ of mobile games, released the beta for its AI engine.

It will enable the next generation of mobile developers to go live with small teams.

In the future, a solopreneur game dev will make a billion dollars - thanks to AI.
Unity CEO John Riccitello knows a thing or two about games.

He has driven the mobile transformation in games the last 17 years.

He said generative AI in games a 10x transformation that will drive exponential growth.

2. AI Antibiotic

Scientists published a paper using AI to discover a new antibiotic.

It came out last week in the prestigious Nature Chemical Biology.

They used deep learning AI methods to prune through many options and find an ideal candidate.

Source: @Nature
AI allows for the rapid exploration of chemical space, increasing the probability of discovering new antibacterial molecules.

The team screened ~7,500 molecules and performed predictions to narrow down to one candidate.

It resulted in publishing in a peer-reviewed journal. Image
3. Minecraft AI

A team including @DrJimFan released Voyager.

It plays Minecraft on its own.

And the crazy part - it just keeps getting better over time 🤯

It writes CODE to play Minecraft.

It's an advanced application of autonomous agents (started before AutoGPT but supported by its reception).

It just keeps improving itself.

And it's totally open-sourced.

4. Picture to 3D

Instorier released a next-level picture to 3d tool.

This is one of the most details and manipulable photo enhancing tools out there.

What a demo.

Instorier is a storytelling tool from photos.

This AI enhancement represents another big leap and shows how the convergence of 3D, Graphics & AI is leading to mind-blowing results.

When focused products add AI, magic happens.
5. MIT: Math AI

MIT researchers released a paper using AI to solve complex math problems much faster than they could have otherwise.

The AI auto-discovers conservation laws from differential equations.
An MIT researcher spent 12 months to solve 12 quantities.

The AI found 14 in less than a minute.

As AI helps us speed up our understanding of basic science - math - the applied stuff will speed up as well. More exponential progress is on the horizon. Image
6. Google Starline

Google has released a prototype of its 3D telepresence product.

The new version uses breakthroughs in AI to render a 3D version of you to make people feel 'present' next to you.

With just 2 unobtrusive cameras.
Thanks to AI, it's progressed very far from the giant booth you needed to use a while ago.

This technology is going to enable people to feel more connected than ever to their loved one's.

And one day, it will land up in Google Meet - to give Google a Zoom differentiator.
7. Nvidia Game AI

Nvidia showcased how AI can power (non-player character) NPC characters life.

It uses Audio-to-speech, AI-powered conversation (think ChatGPT with a voice), and advanced Nvidia graphics tech to create 'next gen' NPCs.
8. Google Flood AI

In its ongoing attempt to "add AI to everything," Google released an AI to forecast flood dangers near your home.

It uses AI to read weather patterns 7 days in advance and predict how deep the waters will get.

And it works GLOBALLY to help those in need.
Best of all - it's completely FREE.

Google Flood Hub is designed to help people who live near rivers and are aware of flooding threats. It gives them access to real-time data to stay safe.

In places like Africa and India where resources are low, even government agencies use it.
9. ChatGPT in Court

A lawyer used ChatGPT in court and didn't check the cases it cited.

Several cases weren't real. ChatGPT made them up.

The lawyer now faces sanctions for his actions.

Source: @verge
The judge wrote a memo calling the circumstance "unprecedented."

The lawyer's brief was filled with citations to non-existent cases.

So, when called out, he went BACK to ChatGPT and ChatGPT doubled down with MORE FAKE QUOTES!

Reality is stranger than fiction. Image
10. TIME: End of Humanity

TIME Magazine joined the bandwagon of AI doomers and questioned if AI could advance so fast that it's become an existential risk.

It published with this cover:

The magazine is worried about the exponential progress of AI.

At some point, it may be hard for us to even keep up with what the AI is doing.

At that point, it could do something seriously upstream of us noticing and cause human extinction.
11. Elon's Neuralink: FDA Approval

Elon is trying to build a high bandwidth connection from the brain to AI.

Our hands are limited to about 10 bits/ second. They're slow. Neuralink could speed that up by a million.

The goal is to supercharge humanity via AI.

Source: @WSJ
Neuralink's vision is to be able to restore vision even for people blind their whole lives.

They've already achieved impressive results in animals.

With the added capacity of human trials with FDA approval, the future is closer than ever. Blind people will be able to see.
12. Few Americans have used ChatGPT Alot

A study by Pew Research found that only 18% of US adults have used ChatGPT a lot.

And there was a huge disparity - only 11% of women.

On the other hand, 42% of American adults haven't even TRIED ChatGPT. Image
Moreover, only 12% of Americans use ChatGPT for work.

We're STILL very early.

88% of Americans have barely scratched their full potential at work with AI. But in 20 years, they all will. Image
13. Japan: No Copyright OK for AI

This last one is special: you won't find it in ANY major news site.

But Japan has affirmed the right of AI companies to train on ANY data.

The controversy with Midjourney is it trained on copyrighted images.

Japan's Midjourney need not fear.
Anime & video game creators are not on board with the decision.

But Japan is making a big push to revive its sluggish innovation the last 15 years.

The country aims to become a world-leader in AI. This decision will help Japanese AI companies get ahead. Image
I hope you found those 26 tweets helpful!

This thread took an enormous amount of time, from reading news to developing the video/image tweets.

Let's have a gentleman's agreement 🤝

1. Sign up for the newsletter (link in bio)
2. Follow me for more @aakashg0
3. Like & retweets

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Aakash Gupta

Aakash Gupta Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @aakashg0

Jul 21
Revealed: How to Succeed in Meta Product Manager Interviews.

Here's the exact doc Meta shares with candidates for its PM Interviews: Image
Image
Image
Read 7 tweets
Jul 1
Steal this template for your homepage: Image
1. Above the Fold

This is the most important part of the homepage. Too many people waste the space:

A. Convey the category of software your product is in, and how your solution differs
B. Add in wow logos from industries in your ICP (ideal customer profile)
2. Three Key Value Props

Remember the rule of threes.

Deliver on your hook above the fold and make your differentiators crystal clear.

Don't just tell, either. Show. Let people experience the product. Attio's videos move when you scroll.
Read 7 tweets
Jun 10
There’s a paradox of seniority: people get more messages, yet they are even more responsive.

This is especially true for important emails.

Here’s why: Image
THE IC LIFE

As an Individual Contributor, your work (outside of certain roles) mostly is internal to your company.

You can spend the vast majority of your day in Slack for responsive tasks.

Email responsiveness matters, but it’s not the most important thing.
THE MIDDLE MANAGER LIFE

As you become a middle manager, email starts to matter more.

You’re hiring folks externally, making calls on agencies… And there’s lots of rapid decision-making to be done cross-functionally.

Responsiveness ramps accordingly.
Read 9 tweets
Jun 1
For a company founded in '93, Nvidia's ascent to $2.7T market cap has been FAST. But what really is Nvidia's moat?

Let's break it down.

PART 1 — SOFTWARE

The story starts all the way back in the early 2000s. That's when Jensen Huang, Nvidia CEO, and his team were out meeting researchers using their products.

Most researchers were hacking graphics packages to run complex parallel compute tasks. It was not ideal. To say the least.

So, when the Nvidia team met Ian Buck, who had the vision of running general purpose programming languages on GPUs, they funded his Ph.D. After graduation, Ian came to Nvidia to commercialize the tech.

Two years later, in 2006, Nvidia released CUDA.

C ompute
U nified
D evice
A rchitecture

CUDA made all those parallelization hacks the researchers were doing available to everyone. Over time, CUDA became the default choice for researchers.

CUDA allowed accessible customization of the low-level hardware. So developers loved it.

Nowadays, when startups like MosaicML evaluate the available technology vs CUDA, they inevitably choose CUDA.

The ecosystem around CUDA has grown so robust that its lead is virtually unbeatable. This software layer is at the core of Nvidia's moat.

PART 2 — HARDWARE

The other side of Nvidia's moat is hardware. But it's not graphics cards for crypto and gaming. The hardware that matters is AI supercomputers.

The story of these supercomputers begins in the late 2000s. As Nvidia was developing CUDA, Jensen asked the team to build a supercomputer to help him build better chips.

The result was a massive supercomputer that weighed 100 pounds and strung together many GPUs with world-class networking for ultra-fast computing.

In the early 2010s, Jensen gave a talk at a conference about this AI supercomputer. Elon Musk got wind of it and said, "I want one."

So, in 2016, Jensen actually donated one to Elon Musk's relatively unknown nonprofit, OpenAI. He hand delivered it, and there's photographic proof.

OpenAI quickly learned the supercomputer worked really well. Especially for training large neural networks. That 2016 Pascal architecture delivered an impressive 19 TFLOPS of FP16 operations.

That's 19 trillion floating point operations per second. It's a massive amount. But that was just the beginning.

Since then, Jensen and the Nvidia team have been lapping the industry in delivering more TFLOPS, growing them at an exponential rate.

The latest Blackwell architecture delivers a massive 5000 TFLOPS. That's >260x AI computer in 8 years. And sells for more than $75K. But buyers like Meta, OpenAI, Google, and Amazon just can't get enough, as their internal ASICs are nowhere near Nvidia's level.

As a result, Nvidia's profits and market cap continue to soar, cementing its position as a leader in the AI hardware and software space.Image
Jensen is one of the most impressive entrepreneurs alive.

He spotted the AI revolution before any other semiconductor CEO and bet the company on it.

That's a rare trait.
And he has a rare management style as well.

He has over 60 direct reports, and doesn't have 1:1s with any of them.

He believes in sharing feedback in public, so everyone can learn from it. And he also does it to remove layers.

Quite the maverick.
Read 5 tweets
May 25
There is no one-size-fits-all when it comes to GTM.

Maja Voje and I studied 12 leading B2B SaaS companies.

(including interviews with their teams)

Here’s what we learned: Image
1. PLG is eating the world

>80% of the companies in our study employ PLG in some fashion.

Even enterprise companies like Snowflake and Salesforce are adding free trials & freemium.

It’s the new normal.
Why is this working for them?

In 2024, the best marketing is often your product.

Users rarely want to lock in a $500K+ contract without trying the product first.

But you do need to layer on a strong product-led sales motion to make enterprise work.
Read 13 tweets
May 16
There's so many ways orgs mess up transforming the product team away from the feature factory.

Here's the top one's, as I see them: Image
1. Understanding and commitment

It's not enough to just have engineering, product, and design on board for transformation.

Transformation impacts marketing, sales, customer success, finance...

You need to drive top-down alignment from the CEO and Senior Leadership Team.
2. Shifting culture

Cultural shift is often the hardest element of a transformation:

• Sales is used to promising features
• Marketing is used to dates when features will arrive

The mistake people make is to gloss over the real tensions for the sake of meeting a schedule.
Read 7 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

:(