- cross-posted to:
- [email protected]
- cross-posted to:
- [email protected]
cross-posted from: https://mander.xyz/post/34629331
cross-posted from: https://programming.dev/post/34472919
cross-posted from: https://mander.xyz/post/34629331
cross-posted from: https://programming.dev/post/34472919
Misleading title. Applies only to AI bought by the Feds.
Watch all the AI companies scramble to comply in a quest for government contracts. This will affect everyone who uses American LLMs and generative AI.
It should also open an opportunity for international competition from less censored models.
And this is one of the best arguments against depending on LLMs. People are outsourcing their thinking to linear algebra machines owned by the wealthy. LLMs are a tool of social control.
Considering how much they bleed cash regularly, I can see them jumping on the government contract bandwagon quickly.
They all got $200M last week
To be fair to the executive order (ugh) many of the examples cited are due to well intentioned system prompts that encourage the LLM to actively be diverse.
The example of a female pope or whatever (read this earlier) is an example of that.
Generally speaking the LLMs have left-bias because they’re trained on information unlike conservatives, but they aren’t necessarily asking the models to be censored
Because Executive Orders aren’t laws. They’re just guidelines for the executive branch of the federal government, which the POTUS is in charge of. It can’t affect private entities like AI businesses, because that would require an actual act of congress.
Notably, this would potentially determine what kinds of contracts the executive branch was able to make. For instance, maybe the government wants to contract out a LLM instead of building their own. This EO could affect which companies are able to bid on that contract, by adding these same restrictions to any LLM that they provide. But on its own, the EO is just that; an order to the executive branch of the federal government.
Then that contracted AI gets used for customer service at a public facing federal agency.
But anything the US feds contracted them for, like building data centres, they have to comply or they face penalties and have to pay all the costs back.
10 days ago, a week before this was announced, they awarded $200M contracts each to Anthropic, OpenAI, Google and xAI
This doesn’t doom the public versions, but they now have a pretty strong incentive to save money and make them comply with the US governments new definition of truth.
Well, in practice, no.
Do you think any corporation is going to bother making a separate model for government contracts versus any other use? I mean, why would they. So unless you can pony up enough cash to compete with a lucrative government contract (and the fact none of us can is, on fact, the while point), the end result will involve these requirements being adopted by the overwhelming majority of generative AI available on the market.
So in reality, no, this absolutely will not be limited to models purchased by the feds. Frankly, I believe choosing to think otherwise to be dangerously naive.
Based on the attempts we’ve seen at censoring AI output so far, there doesn’t seem to me to be a way to actually do this without building a new model with pre-censored training data.
Sure they can tune models, but even “MechaHitler” Grok was still giving some “woke” answers on occasion. I don’t see how this doesn’t either destroy AI’s “usefulness” (not that there’s any usefulness there to begin with) or cost so much to implement that investors pull out because none of the AI companies are profitable, and throwing billions more to sift through and filter the training data pushes profitability even further away (if censoring all the training data is even possible at all).
No. You would use a base model (GPT-4o) to get a reliable language model to which you would add a set of rules that the chat bot follows. Every company has its own rules, it is already widely in use to add data like company-specific manuals and support documents. Not rocketscience at all.
So many examples of this method failing I don’t even know where to start. Most visible, of course, was how that approach failed to stop Grok from “being woke” for like, a year or more.
Frankly, you sound like you’re talking straight out of your ass.
Sure, it can go wrong, it is not fool-proof. Just like building a new model can cause unwanted surprises.
BTW. There are many theories about Grok’s unethical behavior but this one is new to me. The reasons I was familiar with are: unfiltered training data, no ethical output restrictions, programming errors or incorrect system maintenance, strategic errors (Elon!), publishing before proper testing.
why should any llm care about “ethics”?
well obviously it won’t, that’s why you need ethical output restrictions
Honestly this order seems empty. Does the government even have a need for general LLMs? Why would they need an AI to answer simple questions?
As much as I dislike Trump, this shouldn’t impact any AI available to the general public.
Will this stop them from spending our hard earned tax money on it?
They don’t, but they think they do.
to shift blame and responsibility, to create a more modern deity, …
Would you rather our current administration make their decisions by using the lowest bidder LLM, or their own brains?
The LLM probably makes better decisions lol