There is no single current top level. There are multiple disparate levels that have no level above them - the Federal Reserve is one, and Bitcoin is another.
You're missing nepotism from the equation. The salaries available at American Big Tech have made it so that graduates see a dev career as more desirable than traditional prestige industries like investment banking and law. The same nepotistic hiring practices common in those industries will also become entrenched in tech and already are to an extent.
I don't doubt it, but it is also only the past 15-20 years or so when tech became a prestigious career with higher salaries than law and investment banking and not just a "job for nerds". To some extent, the latter is still the reputation it has in the UK/EU where the salaries are nowhere near US levels.
At the minute I would agree that I am going off a broader definition extended to friends and past colleagues but I strongly suspect in the next 10-20 years more traditional family nepotism will seep in as the kids of millennial coders who got their break at FAANG start graduating and using family contacts to land jobs. Yeah engineering will always require a baseline of knowledge but if the basic competence is present there is nothing to suggest that tech is immune to the nepotism which exists in every other industry. The "don't be evil" days are clearly long gone.
Does adding an old book risk making the model worse? Off the top of my head:
- Reinforcing outdated, disproved or otherwise incorrect information.
- Reinforcing outdated forms of communication e.g purple prose.
You could counter both by giving more weight to recent text and I suppose the extra data may help for tracing references and the evolution of ideas through history. If this is what they are resorting to it does feel more like "marginal gains" territory rather than ASI imminent territory
Here's a Ken Livingstone quote from the foreword of "Drama Games: For Those Who Like To Say No":
> In the early 1980s, when the GLC (Greater London Council) was trying to save and create jobs to mitigate the impact of Thatcher's recession, we discovered that the most labour-intensive form of public spending was the arts, and so during the five years from 1981 to 1986 we increased spending on arts and recreation from £l6 million to £160 million. Virtually every actor, painter, poet, sculptor and, in particular, community artist was in work, and it made London a much more exciting city to live in. As well as taking orchestras from the Royal Festival Hall to play in the canteen at Ford's Assembly Plant in Dagenham, we particularly tried to reach disaffected youth. It's against that background that I was able to understand Chris Johnston's book. (Oh, and by the way, if you want to know which is the least labour-intensive form of public spending, it is the military.)
Maybe if we stopped pushing insane amounts of money into fossil fuels and the military industrial complex, and instead redirected it into the arts and sciences then, just maybe, we might actually end up with a happier, more employed, more fulfilled, and more equal planet.
This only works if you can force every state to do it, lest naive states be defenseless against those that don't disarm. But any such compulsion would be self-terminating, and thus unstable.
Why not just do it all? Fund the military and the arts and the sciences.
The thing is what is the alternative option? If you don't build it, then your rivals do. All of your electronic and military systems then become compromised because your rivals can exploit or negate them one way or another via AI. So you have to build it regardless. It's a suicide pact the same way nuclear weapons are a suicide pact but if you don't have them then you're at the mercy of anyone else who does.
Build it without the catastrophising PR. Try, you know, selling it as something that will do good.
Altman wanking on about how he’s going to end the species is great for attracting investors—this crap worked for Bankman-Fried, too [1]. If you are building world-ending kit, you’re a super-important dude who has to be listened to. On the other end of the spectrum, you have everyone quietly doing their work.
Sometimes I wonder if what will actually happen, is that we will have very small startups, essentially a C-Suite but with actual coding skills and domain knowledge, who build new companies using AI that end up gaining an advantage over the incumbents and force them out of business. Indirectly you're replaced by AI. I could see the drasticity of the changes to culture and workflow that AI demands being too much for most legacy companies to handle. For instance, there are still a shockingly large number of companies relying on seriously dated software and paper based systems.
reply