It’s a Huge Mistake to Discount GPTs and LLMs (AI) Due to the Crypto Bubble…
Feeling the crypto bubble hangover?
Don’t let it make you sleep on GPT4.
It’s essential not to overlook the significant advances in GPTs (Generative Pre-trained Transformers) and Large Language Models since GPT-3’s release in November 2022.
My qualifications….
I worked in an ‘emerging tech’ publishing research firm covering crypto for a while. I’m no stranger to bubbles.
I have been a crypto user since 2014 and publicly called the TOP of the 2021 crypto cycle down to the exact day in Nov 10, 2021 as a speculator — and I remain in the industry as a native technology lover and crypto user.
How does the AI and blockchain hype cycle compare?
What’s behind the differences in the adoption and hype in wildly different macro environments?
Is AI more captivating to us psychologically?
Differences in broad-based applicability and speed to mass market?
Maybe something re: the #singularity or direct human augmentation?
This is not to disparage crypto nor compare two different technologies, but to highlight some differences in the speculation.
I’m not claiming that AI is immune to hype or bubble dynamics. As with any new tech, expected future values can make present-day valuations appear inflated.
And one day, there will be an AI hangover as well.
The Key Differences between the crypto bubble vs the AI bubble
(not between the TECH)…
- Use case vs. speculation — due to broad applicability for AI and therefore speed to mass market vs. more narrow or infrastructure/tooling applicability for blockchain
- Real economy adoption vs. isolated crypto-financial economy adoption
- Resulting real and perceived differences in productivity / efficiency gains
- Happening in wildly different macro environments
- Expectations vs. reality
Adoption
Generative AI is already being used in wide ranging industries — mostly for creative and professional work involving content generation, data and information processing, and even higher level strategizing and planning.
GPT4 surprises me daily.
Blockchain applications have mostly been perceived as being limited in scope to trading and crypto-financial use cases — but this has changed, especially in 2022 with major non-financial brands getting into NFTs, for example.
And other notable exceptions in government and enterprise use cases (supply chain, police and education records on Polygon blockchain, tokenized private equity funds and even real estate, DAOs).
Recent research by UPenn — read the abstract: https://arxiv.org/pdf/2303.10130.pdf
“Our analysis suggests that, with access to an LLM, about 15% of all worker tasks in the US could be completed significantly faster at the same level of quality. When incorporating software and tooling built on top of LLMs, this share increases to between 47 and 56% of all tasks.
This finding implies that LLM-powered software will have a substantial effect on scaling the economic impacts of the underlying models.
We conclude that LLMs such as GPTs exhibit traits of general-purpose technologies, indicating that they could have considerable economic, social, and policy implications.”
Productivity and Efficiency
The SMB market is rapidly adopting AI for both business and personal use.
Emad Mostaque of Stability AI notes that F500 companies have not yet been able to do the same due to compliance and data privacy reasons.
But this is shifting fast with the rise of private LLMs.
That’s not to say that blockchain innovation has not created productivity and efficiency gains.
Instant, low fee cross-border transfers at any time or day were revolutionary.
DeFi saw the rise of open-access global financial markets offering lending and other financial services for anyone in the world without the need for any financial intermediaries.
However, the broader applicability of AI has had a more direct impact on augmenting human capabilities and accelerating productivity across various sectors — and therefore, excitement.
Don’t get me wrong… DeFi summer was exciting. Being able to farm food tokens and pledge crypto as collateral was neat.
But it didn’t provoke existential anxiety the same way.
The Masses’ Reaction
Most notably — the masses are trying to use AI, not speculate.
GPTs have already exceeded expectations in their current form. Not to overhype GPT4, but it truly looks to be at least as important to learn to properly use and prompt as using Google was in the 2000s.
The AI hype notably remains strong in a 5% rates environment with macro headwinds, and we’re seeing the proliferation of AI services.
Crypto asset valuations are struggling in the post-ZIRP environment so far, but dedicated builders keep building.
Another litmus test that shows the scope of the tech so far…
Crypto regulations primarily deal with securities laws, money laundering, and hacks, while AI regulations, though still largely absent, focus on wide-ranging, unimaginable use cases and existential concerns.
This is of course due to the reality that much of crypto use has so far been around speculating and in financial markets.
Conclusion
In conclusion, referring to the broader applicability thus far, AI has had a wider reaching impact on accelerating productivity for the everyday person via directly augmenting human capabilities, in very real terms that apply to literally everyone.
In ways that the public did not expect even just 6 months ago — so expectations vs. reality is also a factor at play.
The barrier to entry and UX is also vastly different for GPT — and the risks — so far — are minimal unless you’re letting a rogue autonomous GPT agent run costly API calls and browse the internet on its own.
The promise of decentralized ecosystems and innovations we’ve seen over the last year in blockchain tech are undeniably valuable.
But the speed to mass market and broad adoption in general terms is what’s caught everyone off guard.
As for crypto… the EU just passed MiCA regulations this week so hopefully we see continued innovation and stability for the industry after a tumultuous 2022.
And hopefully one day the promise of blockchain innovation will too be comparable in scope to what we have seen in a short 6 months (really years in the making) from GPTs.
Of course, we are yet living in a sci-fi post-scarcity technocratic utopia just yet.
