Among other reasons, if you turn temperature down to 0, llms stop working. Like they don't give natural language answers to natural language questions any more, they just halt immediately. Temperature gives the model wiggle room to emit something plausible sounding rather than clam up when presented an input that wasn't verbatim in the training data (such as the system prompt).
I don't think so. Primarily because if you can ask that question instead of just being dead, then it's not the fast takeoff.
On a less drastic note, if AI were autonomously cannibalizing the economy, you'd run into more things and go "huh I guess that's run by AI now" instead of "ah godammit why did they shove an LLM in this workflow".
I don't think adding 20 bytes to every response, which are never read in practice, will improve the energy efficiency of the internet. Not to be mean, there's just a lot of stuff in most responses that effectively nothing acts on.
We're not it's competition in the same way chimpanzees aren't our competition. Some fraction of us are interested in their well-being for aesthetic reasons, but a lot of the time this fraction loses to a not particularly powerful faction in direct competition for territory. And if there is any serious conflict of interest, there is no contest and the chimps lose. If we get lucky some fraction of superintelligence will look on us the way we look at ground apes, but that's far from a given.
I don't think it's right that Airbnb solved short term rentals - outside of a few dozens prestige markets that they monitor with humans, it's really a race to the bottom. Reputation at scale remains unsolved, and so it's still a market for lemons.
You could get healthcare from someone who doesn't have to go $100ks in debt first, and then expects to make more than you do. If no such person with that profile exists, then yeah you're stuck paying the rate of the system that trained and employs them.
At the end of the day, much like housing cannot be both affordable and a good investment, healthcare cannot be both affordable and a lucrative career. Avoid going to an MD if you can get the same care from someone else.
Kanban operates on a known product with a known manufacturing process. Many software products are undefined even at time of public release, and evolve continuously. "Deciding what to build", while explicitly highlighted in the agile software manifesto, is the weak link.
Put another way, lean manufacturing improves metrics for the margianal unit of goods. No one, customer or dev, is interested in the margianal unit of software.
I would guess not. An important feature of compilers is that they are guaranteed to emit code with certain properties in response to specific inputs (memory safety guarantees, asymptotic performance, calling convention, etc.). If they don't do that, you can file a bug report.
You cannot file a bug report against an LLM that it produced an unexpected output, because there is no expected output; The core feature of an LLM is that neither you nor the LLM developer knows what it will output for a wide range of inputs. I think there are a wide range of applications for which LLMs core value proposition of "no-one knows a priori what this tool will emit" is disqualifying.
Audio and video are encoded, compressed, and transmitted differently because of how humans audially/visually decode them. We fare better dropping late video frames than degrading their quality, where the opposite is true for audio. As we transmit the two as separate, asynchronous signals, it's unsurprising that they are frequently out of sync.
If you're concerned that you've made/found something dangerous, the most appropriate solution is to disclose it to people who can evaluate how dangerous it is, and work with them on next steps.
In your hypothetical, I would observe that Paul Christiano is close to being the person in the US government in charge of evaluating AI danger, and that he is currently involved with https://www.alignment.org/ who specifically consider issues of AI risk. I would humbly contact them and ask for guidance on how to proceed.
Be prepared to be dismissed as a crank. Remember that avoiding harm to others is more important than proving that you're right.
Not significantly. Humans are made up of matter already available at the earth's surface, so population increase alone effects neither the amount nor distribution of the earth's mass.
Technological civilization might at some point meaningfully shift the distribution of mass, but I don't think it has up to this point.
You ought to begin by getting really good citations on each of those figures. If you or the source you got them from had any confusion about how to arrive at those numbers, any results you get from them will likely be meaningless.
Hopefully in doing so you will start to notice the missing pieces of info and can ask more targeted questions to get better results.
People are more impressed by things they cannot do than by things they can. The vast majority of people in the industrialized world are functional writers; the portion who are competent performers is much lower.
While professional writers may be quite skilled, the gap between what they do and what the median adult does seems traversal. The psychic distance is much larger for other creative endeavors.
Focus on something you expect AI not to exceed professional humans at in the next couple of decades. If you're convinced there's no such thing, then I wouldn't worry so much about career choice.
For a concrete answer in tech, have a look at AI development itself. Keep abreast of what OpenAI and their competitors identify as open problems and orient your academic career towards working in them. The object level questions will change while you're in school; the goal is an "intercept trajectory" where what you study just before you graduate is the state of the art.
Alternatively, highly regulated professions (medicine, law, civil engineering to name a few) are likely to continue to employ bright humans long after AI can do the job just because no-one will be allowed to use AI there.