I think taking Anthropic or any company in this space at face value is naive at best though. AGI has been 6 months away for years now. Surely anyone can think this through: Anthropic knows what theyre doing with their public facing repositories, they know to make things enabled by their tech seem impressive. I would consider Bun etc. examples of this.
The disagreement in timelines usually comes from differences in the used definition of AGI.
Many who predict a 2026–2028 arrival define it as: 'the ability to perform any purely cognitive task a human can do.'
If we stick strictly to that 'cognitive-only' metric, we are arguably very close.
Plenty of people say vaccines contain microchips. Research is not linear so this simple linear regression I see people doing regarding capability scaling makes no sense. Nobody can forecast a breakthrough, what makes you believe you can?
I ahven't seen technology like this which affects me directly. But from past we know what disruptive technology looks and feels like. The weaving chair/loom disrupted an industry, energy, steam engine, internet.
Now we have the most generic technology an LLM/AI at a time were every major problem was solved:
We have the internet which allows for fast communication, we have very fast hardware, we have a supply chain which can react/act very fast globally and we have the richest companies on our planet investing into this technology unseen amounts of money. We have a local race between the richest companies on the world and a global race between the biggest world powers on our planet.
And we have the smartest people on our planet involved in this too. A lot of peple from academy went to the industry, some gave up their tenure for AI.
It would be very ignorant to assume that all of this can't lead to significant and fast change in our society.
It might not, but its like looking at a huge wildfire 50km away and just going to bed.
Realistically, nobody intellectually honest really knows.