As a part of Meta’s Q3 earnings report final week, CEO Mark Zuckerberg made a quick observe that raised many eyebrows in response.
Within the earnings name, Zuckerberg famous that:
“This quarter, we launched Llama 3.2, together with the main small fashions that run on-device and open supply multi-modal fashions. We’re working with enterprises to make it simpler to make use of, and now we’re additionally working with the general public sector to undertake Llama throughout the US authorities.”
Meta’s working with the federal government on the adoption of AI in authorities functions? What might that imply?
Properly, in the present day, Meta has supplied extra context, with its head of Authorities Affairs Nick Clegg explaining that Meta is certainly working with U.S. Authorities companies on potential functions for its Llama AI fashions, whereas additionally partnering with some big-name companies on the identical.
As defined by Clegg:
“Meta’s open supply Llama fashions are more and more being utilized by a broad group of researchers, entrepreneurs, builders and authorities our bodies. We’re happy to substantiate that we’re additionally making Llama obtainable to U.S. authorities companies, together with these which might be engaged on protection and nationwide safety functions, and personal sector companions supporting their work. We’re partnering with corporations together with Accenture, Amazon Net Companies, Anduril, Booz Allen, Databricks, Deloitte, IBM, Leidos, Lockheed Martin, Microsoft, Oracle, Palantir, Scale AI and Snowflake to deliver Llama to authorities companies.”
So the generative AI methods that commonly misread information from the online, and convert them into convoluted, unreal narratives, at the moment are going to be powering parts for the Protection Division. Bought it, sounds good.
Nah, it needs to be nice. I imply, it’s not like they’re going to make use of it for mission planning or something main like that.
“Scale AI is fine-tuning Llama to assist particular nationwide safety staff missions, resembling planning operations and figuring out adversaries’ vulnerabilities.”
Oh.
I imply, that looks as if a priority.
General, nonetheless, Clegg says that Meta’s AI fashions are being examined in functions the place they will simplify knowledge evaluation and interpretation.
“[Large language models] can assist to streamline sophisticated logistics and planning, monitor terrorist financing or strengthen our cyber defenses. For many years, open supply methods have been crucial to serving to the USA construct essentially the most technologically superior navy on the earth and, in partnership with its allies, develop world requirements for brand spanking new expertise.”
The adoption of the newest AI fashions, then, is in keeping with holding the U.S. Authorities a step forward of potential enemies, and utilizing the very best instruments obtainable to enhance its operations.
So it is sensible. Nothing to be involved about.
I imply, it’s not like they’re going to develop robots and provides over management of them to AI methods and people methods are then going to take over and decide that people are the true enemy and eradicate us all.
“In July, a former chairman of the Joint Chiefs of Workers mentioned that robots and different good equipment will comprise as much as one-third of the U.S. navy within the subsequent 10-15 years.”
…
Okay then.
Properly, I imply, all of us have sufficient to be pressured about this week, so perhaps don’t suppose too arduous about Meta giving over AI fashions to the U.S. Protection Division, which can also be creating autonomous robots outfitted for the battlefield.
