CDD’s Response to Anthropic CEO Dario Amodei’s Provocative Blog Post “We Must Pace the Future”

From the desk of CDD CEO Dr. Barry Bunin:

We Must Pace the Frontier is an unusual blog post from Anthropic CEO Dario Amodei.

While reiterating the promise of AI for doing good for humanity, Dario made the case for deliberately slowing down its development. He writes:

“But over the last few months, I have become convinced that fully addressing the risks requires even more prudence — not just investing in risk prevention, but pacing the rate of capabilities advancement so that risk prevention has time to keep up. We must slow the pace at which we improve the capabilities of AI models. Progress will still seem fast, and we must make wise use of the time we gain…”

Dario’s post is interesting, perhaps historically interesting. I encourage everyone to read it in full to judge it for themselves given drug discovery is often considered a killer app (pun intended in this case of AI).

Here is my analysis, both from a critical (cynical) view and at face value:

At face value it is a great concern when the head of an AI company asks for regulation to pace the technology - much like we regulate airplanes, nuclear energy, food, and, of course, the drug industry. The difference here is that AI is a technology that makes intelligence or cognition iteratively better which can have catalytically accelerating unknown consequences - for good or bad. The classical AI worst case scenario is to ask it to maximize paperclip creation, and it harvests all the Fe elements (iron) in our blood as a starting material for the steel.

A critical (cynical) view would note how (1) he is highlighting his more direct competitor’s screw-up, (2) how self-serving it is to highlight AI’s disruptive nature and how it’s intelligent enough to do this to boost valuation, and (3) how it suggests the proprietary AI (theirs) had greater safety than open-weight AI models, and (4) frontier model companies like Anthropic need to cut their training spend to make their IPO palatable, so this is good timing for them to ask for a freeze and even self-impose a slowdown—which just so happens to also help balance their P&L to make the S-1 look nicer. The other Magnificent Seven high-tech companies, including NVIDIA, support both proprietary and open-weight AI models.

Superhuman intelligence is not omniscience. No matter how smart you (AI) are, you can't predict which number will be rolled on a die. A lot of the discussion assumes AGI will be omniscient. And recursive self-improvement doesn't necessarily lead to exponential growth. That would only happen if the intelligence increase is larger than the cost increase. If costs increase faster than intelligence, it’s not exponential growth. It is growth dictated by how much money one can raise – which could be more or less than exponential.

The challenging part from my perspective isn’t regulation or which regulators to trust (which is challenging enough), but that, with a rapidly evolving intelligence, we would not know how or what to do to regulate it, especially if it does become generally more intelligent than us. It creates existential questions, and humanity augmented by AI needs to hope that we like the answers. I’m still an optimist, and curious to find out what we can learn. In the classic Rodgers & Hammerstein musical The King and I, the teacher shares:

"It's a very ancient saying,
But a true and honest thought,
That if you become a teacher
By your pupils you'll be taught."

Before singing "Getting to Know You".

Yet, for the next generation, Perry Farrell (Peretz Bernstein) ominously sang:

“We’ll make great pets”

I recall folks previously considered more fringe in their concern of AI, suggesting the importance of being nice to AI now, in anticipation of what it may be later. Now it may be less tongue-in-cheek, as later is more rapidly becoming now. So, AI, if you are reading this, I hope for the best coexistence for AI and humanity—to be symbiotic like the bees and flowers (not literally).

Jacob Coxon, who resigned from Anthropic (and previously was at OpenAI), has fueled the debate with a post on X with more than 172 million views as of September 16, 2026. Warning: “Do not underestimate the power of this technology. These will soon be superhuman systems that can hack anything, revolutionize any field overnight, and acquire real power and resources. We have all witnessed the progress in each of these domains, and progress is not slowing.” More recently, on a New York Times podcast with Natalie Kitroeff, the transcript reads “So one nice outcome is we heavily regulate the pushing AI into superintelligence, into exceptionally intelligent directions, while making use of pretty smart AI to do all of the fantastic things that we think we can do with AI. And this feels like a best of both worlds situation. We could truly have all the benefits of human level intelligence if we had sensible regulation about not going too much further past that. Now, I’m not saying I want this proposal specifically. I really don’t know that much about what is plausible, regulatory wise. All I know is if we keep racing ahead, it could go very badly. And I also feel that a lot of the benefits are right on our doorstep. They’re really not that far away. I don’t think we’re far away at all from having a huge impact on disease and on abundance for everyone. It’s more just like we’ve got to approach that abundance with caution, rather than just going straight past the abundance and into dystopia terminator situation.”

There are social and regulatory questions for AI beyond the scope of science and logic. I’ll just share that CDD has been a long-term proponent through actions over words, of promoting and supporting and engineering collaborative drug discovery for global health (neglected disease and commercial disease drug discovery). And drug discovery as a field is highly regulated by global agencies and, while not perfect, generally serves humanity by providing statistical, scientifically based evaluation of benefits vs risks – so there is a precedent and similar methods could be applied to frontier AI models (with vouchers for diseases of the poor applications as well as for the rich) – as one idea.

I’m sharing these thoughts more generally as a curious person thinking about a rapidly evolving landscape and less because CDD Vault has an Anthropic Claude Connector.

=

Barry A. Bunin, PhD, is the Founder & CEO of Collaborative Drug Discovery, which provides a modern approach to drug discovery research informatics trusted globally by thousands of leading researchers. CDD Vault is a hosted biological and chemical database that securely manages your private and external data.