State of AI: Why Ozempic and Blockchain Are Both Incentives to Develop AI Models

I know, there is no real need to move faster on AI progress than it already is.

But the reality is other technologies are ‘converging’ to AI: drug discovery and blockchain.

-Drug discovery.

Since the first version of AlphaFold 7 years ago (yes that long), drug discovery has been on the radar as a natural AI application.?

However, FDA drug approval still remained a daunting process and the best (so far) application of the technology was a ‘variation’ of the same molecule (briefly Alphafold can experiment faster with different ‘orientations’ of the molecule with various permutations taken into account).

But then Covid came along. And now GLP-1 drugs (Ozempic and Wegovy for now) came along.?

What’s the matter? Well Covid allowed a ‘faster’ route to get FDA approval and GLP-1 drugs showed potential beyond weight loss (like 40% of the US population is affected)… you just need to find the ‘best orientation’...

Of course I am simplifying but you should get the gist.

-Blockchain.

Beyond Bitcoin and few other cryptos, it is hard to find an easily understandable application of the technology.

However in AI, you need now to train massive models that require huge computational power. Instead of having a dedicated data center to compute, an idea is to share the computation among different users across the internet. More or less like torrent works for sharing files. Now, in the case of training AI models, sharing AI models or just sharing the data for the AI models, it costs money for the user.

It is not like using torrent, where the user has minimal costs to sustain sharing info with others. Having a GPU crunching data costs money. Hence you need an incentive… nothing better to have a blockchain with a native token to reward the users who contribute with the most computing power, the most (or better) data, the better model. We are seeing the emergence of solutions like Render, Bittensor, Virtuals etc. all incentivising users to share their own ‘contribution to AI’ in exchange for tokens.

And these are just a few examples of what AI ‘convergence’ with other technologies or industries can enroll in front of us.

If you want a full in depth discussion on all the possible topics ‘colliding to AI’, I have hosted a live event State of AI 2025 report: The Full Deep Dive Part 1 (no matter how long)’

In this event, and in the Part 1one (February 26th), https://www.dhirubhai.net/events/stateofai2025report-thefulldeep7297703741493829633/theater/,?

I try my best to exhaust all the possible angles where AI can make an impact in the next 1-2 years.?

I hope it helps.

#ai #artificialintelligence #business #technology? #innovation #hardware #law #science #crypto