Hacker News: adamsmith143

New comment by adamsmith143 in "Discovery of spherules of likely extrasolar composition in the Pacific Ocean [pdf]"

adamsmith143 — Wed, 30 Aug 2023 15:56:41 +0000

Are you also a professional Astronomer that can accurately vet his claims and research?

New comment by adamsmith143 in "AI demand is already shrinking"

adamsmith143 — Wed, 23 Aug 2023 18:19:06 +0000

My own observation while using the OpenAI API is that it is massively rate limited due to the enormous demand. So while the public hype may have lessened the actual demand for their core product is so large that its problematic for users.

New comment by adamsmith143 in "Northvolt raises $1.2B convertible note from BlackRock"

adamsmith143 — Tue, 22 Aug 2023 15:57:32 +0000

Generally have been satisfied with Anker, though that stuff is made in China the chargers and batteries have been good. Monoprice used to be good but I haven't used anything from them in a while

New comment by adamsmith143 in "GPT-4 can't reason"

adamsmith143 — Tue, 08 Aug 2023 20:29:49 +0000

You should probably think about why you think that making the model output reasoning steps which lead it to correctly answer questions it couldn't before is not somehow equivalent to reasoning.

New comment by adamsmith143 in "GPT-4 can't reason"

adamsmith143 — Tue, 08 Aug 2023 16:10:29 +0000

Yeah man they have teams on standby to adjust the model whenever a random unknown author posts something on obscure pre-print servers. Then they spend hundreds of thousands of compute $ to improve the model on that one metric the paper attacks.

New comment by adamsmith143 in "GPT-4 can't reason"

adamsmith143 — Tue, 08 Aug 2023 16:03:36 +0000

>You had time to respond though, what a silly (and elitist rebuttal).

What a dumb take. It probably takes seconds to write a simple comment and far longer to read a paper.

>You argue by authority that yes, it actually does reason, which is a far (far) bolder claim than the one OP is making.

He linked a paper you can read yourself. How is this arguing from authority?

New comment by adamsmith143 in "GPT-4 can't reason"

adamsmith143 — Tue, 08 Aug 2023 15:58:38 +0000

The probability that a Stochastic Parrot returns coherent reasoning seems vanishingly small.

New comment by adamsmith143 in "GPT-4 can't reason"

adamsmith143 — Tue, 08 Aug 2023 15:55:06 +0000

"Stochastic Parrot" is a really tired take and none of the major players, Ilya Sutskever, Andrej Karpathy, etc. believe that's all these models are doing.

New comment by adamsmith143 in "How real is America’s chipmaking renaissance?"

adamsmith143 — Tue, 08 Aug 2023 15:45:50 +0000

Quartz = SiO2 --> Silicon and Oxygen QED

New comment by adamsmith143 in "Observation of zero resistance above 100 K in Pb₁₀₋ₓCuₓ(PO₄)₆O"

adamsmith143 — Thu, 03 Aug 2023 14:32:42 +0000

With you there. A lot of Chinese labs are listed e.g. but my own experience trying to replicate materials science results from "credible" Chinese labs has led to a lot of disappointment.

New comment by adamsmith143 in "Observation of zero resistance above 100 K in Pb₁₀₋ₓCuₓ(PO₄)₆O"

adamsmith143 — Thu, 03 Aug 2023 14:31:06 +0000

The proposed mechanism for superconductivity in lk-99 is pretty different so that may be confounding your interpretation of those graphs.

New comment by adamsmith143 in "Twitter’s Rebrand to X Could Be a Trademark Nightmare Thanks to Microsoft"

adamsmith143 — Tue, 25 Jul 2023 14:36:00 +0000

Yeah so hard to make money during the Dotcom boom with all your dad's apartheid emerald mine money.

How does his boot taste?

New comment by adamsmith143 in "Twitter’s Rebrand to X Could Be a Trademark Nightmare Thanks to Microsoft"

adamsmith143 — Tue, 25 Jul 2023 13:59:43 +0000

What?? You mean Elon made another colossal mistake, bbbut hes such a business genius...

New comment by adamsmith143 in "The economics of the Birkin bag (2016)"

adamsmith143 — Mon, 24 Jul 2023 20:36:29 +0000

>Its interesting how many lower income people have iphones, Jordans, Teslas/Mustangs/Corvettes

Yeah all those food stampers buying Corvettes and Teslas is so annoying. Lolwut

New comment by adamsmith143 in "AI researcher Geoffrey Hinton thinks AI has or will have emotions"

adamsmith143 — Mon, 24 Jul 2023 17:25:54 +0000

This is of course false. We can't prove how any system is able to have emotions much less prove that a system can't

New comment by adamsmith143 in "AI researcher Geoffrey Hinton thinks AI has or will have emotions"

adamsmith143 — Mon, 24 Jul 2023 17:23:06 +0000

You jest but that is actually a subject of real research.

New comment by adamsmith143 in "AI researcher Geoffrey Hinton thinks AI has or will have emotions"

adamsmith143 — Mon, 24 Jul 2023 17:21:20 +0000

The same can be said about meat brains, they just execute natural programs. There's no reason in principle why a meat brain should be able to experience emotion and silicon brains cannot.

New comment by adamsmith143 in "White House Secures Commitments from Leading AI Companies to Manage AI Risks"

adamsmith143 — Fri, 21 Jul 2023 20:25:25 +0000

Ah yes all that research money will be used to buy mansions and ferarris. Do you have a clue how this stuff works in real life?

Also note that until maybe the last year not only was there not a great financial incentive to be a "doomer" but it would actively hurt your career if you were. Most of the main people in the scene have been sounding this alarm for years sometimes decades. It's also difficult to explain people like Geoff Hinton or Yoshua Bengio joining and leaving behind high profile and highly lucrative positions. Yann LeCun, a staunch anti-doomer is perhaps the perfect example who has actually DOES have enormous financial incentive to play down AI dangers.

New comment by adamsmith143 in "White House Secures Commitments from Leading AI Companies to Manage AI Risks"

adamsmith143 — Fri, 21 Jul 2023 20:24:36 +0000

So the fact that thousands of researchers with billions of dollars behind them to build this omnipotent being means nothing to you? Because they all certainly think it's achievable.

New comment by adamsmith143 in "White House Secures Commitments from Leading AI Companies to Manage AI Risks"

adamsmith143 — Fri, 21 Jul 2023 14:16:37 +0000

If you look at actual quotes from people like Sam Altman, Ilya Sutskever, Demis Hassabis, etc. the AI, that is to say really the AGI, that they want to build is a godlike super intelligent being. To many researchers in the field this is their end goal.