AI Project Hub (Ø,G) @AIprojecthub - Twitter Profile

Pinned Tweet

over 1 year ago

Why will a decentralized open-source AI protocol like Bittensor win? > Innovation - Look at Deepseek R1...96.4% cheaper than OpenAI o1 while delivering similar performance. Again...96.4% cheaper. AI should be accessible to those in all countries and many cannot shell out $200 a month for OpenAI. Within hours, it was on Bittensor. (Play with it on Chutes - Powered by Bittensor - https://t.co/hARuYqEceR) > Results - Countless Subnet teams have already outperformed their closed-source counterparts on benchmarks pertaining to their specialty. This isn't a theory. It's already happened. The power of incentives is mind-blowing. Source: Top miners outcometing Zestimate ( https://t.co/XBQv3bwwYU) Source: Detecting Deepfakes at 90% (world leading) - https://t.co/ChmB4LKudK Need more accomplishments? Check the attached image (note: 4 months old) I highly respect @macrocrux and in a recent interview, he mentioned his PHd ML engineering staff would have trouble competing with some miners. That is similar to winning a Kaggle competition, being a top miner is Bittensor is the new bragging rights. It's *that* competitive. (Here's what his team has been up to - https://t.co/jR12WS2r2K) > Worldwide Talent - Imagine anyone with the skills to make something better could contribute. Anyone. Doesn't matter where they are. Results matter. Opensource allows all to play without going through an HR department, background checks, agreeing to company policies, missing your kids grow up, etc. Even huge budgets won't cut it to hire the best. Some have personal, political, or regional reasons; they don't want a typical job. > Decentralization - Risk mitigation...i.e - Seems much easier to launch a cyber attack on a known location of GPUs as opposed to a decentralized network running all over the world. For something as important as AI, seems like a critically important design feature to consider? If major governments gets hacked on average nearly once a month, if someone wanted to attack the holygrail of AI, what would stop them? Seems easier to have a distributed network running across the world. (source: https://t.co/gamnF1l0PX.) > Permisionless - with $dTAO on the horizon, hundreds and eventually thousands of subnets will be on the network without a huge barrier of entry. Pushing the boundaries yet again, with elite miners outcompeting their closed-source counterparts as already evidenced. Sometimes in as quickly as a few months. These are my quick thoughts on how Bittensor and $TAO will lead the way. It isn't theoretical. It's already been set in motion. Nothing can stop it. Just like the financial system couldn't stop #bitcoin.

AIprojecthub's tweet photo. Why will a decentralized open-source AI protocol like Bittensor win?

> Innovation - Look at Deepseek R1...96.4% cheaper than OpenAI o1 while delivering similar performance. Again...96.4% cheaper. AI should be accessible to those in all countries and many cannot shell out $200 a month for OpenAI. Within hours, it was on Bittensor.

(Play with it on Chutes - Powered by Bittensor - https://t.co/hARuYqEceR)

> Results - Countless Subnet teams have already outperformed their closed-source counterparts on benchmarks pertaining to their specialty. This isn't a theory. It's already happened. The power of incentives is mind-blowing.

Source: Top miners outcometing Zestimate ( https://t.co/XBQv3bwwYU)

Source: Detecting Deepfakes at 90% (world leading) - https://t.co/ChmB4LKudK

Need more accomplishments? Check the attached image (note: 4 months old)

I highly respect @macrocrux and in a recent interview, he mentioned his PHd ML engineering staff would have trouble competing with some miners. That is similar to winning a Kaggle competition, being a top miner is Bittensor is the new bragging rights. It's *that* competitive.

(Here's what his team has been up to - https://t.co/jR12WS2r2K)

> Worldwide Talent - Imagine anyone with the skills to make something better could contribute. Anyone. Doesn't matter where they are. Results matter. Opensource allows all to play without going through an HR department, background checks, agreeing to company policies, missing your kids grow up, etc.

Even huge budgets won't cut it to hire the best. Some have personal, political, or regional reasons; they don't want a typical job.

> Decentralization - Risk mitigation...i.e - Seems much easier to launch a cyber attack on a known location of GPUs as opposed to a decentralized network running all over the world. For something as important as AI, seems like a critically important design feature to consider?

If major governments gets hacked on average nearly once a month, if someone wanted to attack the holygrail of AI, what would stop them? Seems easier to have a distributed network running across the world.

(source: https://t.co/gamnF1l0PX.)

> Permisionless - with $dTAO on the horizon, hundreds and eventually thousands of subnets will be on the network without a huge barrier of entry. Pushing the boundaries yet again, with elite miners outcompeting their closed-source counterparts as already evidenced. Sometimes in as quickly as a few months.

These are my quick thoughts on how Bittensor and $TAO will lead the way. It isn't theoretical. It's already been set in motion. Nothing can stop it. Just like the financial system couldn't stop #bitcoin.

Chamath Palihapitiya

@chamath

over 1 year ago

Let me say the quiet part out loud: AI model building is a money trap. What you are seeing here is a feature of modern AI models: there is no bounding law like Moore’s law where advances can be predictably expected. Nor has any model builder adequately publicly explained what to expect sans jargon and BS. What this means is that it’s very difficult to plan spending programs to create advances and will likely render entire months and possibly years of energy, hardware, R&D, OpEx and CapEx wasted and useless as new innovations are discovered. Open Source is the clear winner. Closed source will be forced to keep their best models secret and sell to enterprises OR try and create some incredible consumer app with it. Having done this before, my lived experience is that this is easier said than done. Net/net consumers will win, many venture funded efforts will not.

150

3K

361

1K

1M

6

77

22

13

16K

AI Project Hub (Ø,G)

@AIprojecthub

Who to follow

Last Seen Users on Sotwe

Trends for you

Most Popular Users