Opens profile photo
Follow
Harold Godsoe
@HaroldGodsoe
Canadian cross-border U.S. corporate lawyer in Japan
Lawyer & Law FirmTokyo, JapanJoined December 2011

Harold Godsoe’s Tweets

why are people always trying to change the topic from "which propositions should we believe" to "which feelings should we feel"
Quote Tweet
"anxieties are understandable, but … much overblown. I’ve heard similar fears related to the advent of Artificial General Intelligence expressed off and on since I introduced the term AGI in 2005, but I think a pause would be a badly wrong move in the current situation" twitter.com/bengoertzel/st…
7
116
Hinton is one of the greatest AI scientists to ever live, and he has quit Google in order to talk about the dangers of AI freely. In case anyone was still somehow on the fence about whether things are serious.
Quote Tweet
“The idea that this stuff could actually get smarter than people — a few people believed that. But most people thought it was way off. And I thought it was way off. I thought it was 30 to 50 years or even longer away. Obviously, I no longer think that.” nytimes.com/2023/05/01/tec
Show this thread
13
489
We're thrilled to see Members of the respond to FLI's open letter! A group of 12 MEP's led by and are calling for: - Tailored rules for foundation models in the EU AI Act - A high level global summit on AI - Democratic oversight and… Show more
Image
Image
Quote Tweet
Image
Image
AI is moving very fast and we need to move too. The call from the Future of Life Institute @FLIxrisk to pause the development of very powerful AI for half a year, although unnecessarily alarmist, is another signal we need to focus serious political attention on this issue.
Show this thread
3
68
This is a nice demonstration of the old Yudkowsky argument of how even linear improvement can look like a dramatic transition when the human ability range is fairly narrow. I predict it will however be more revolutionary in supplying problems than solving them.
Quote Tweet
New paper comparing GPT-3.5 & GPT-4 performance on college physics problems. It shows that in just a few months, AI has made a leap from the 39th to the 96th percentile of human level performance. Now imagine were it will be in 10 years.
Image
15
260
Show this thread
I'd strongly support the idea of a Manhattan Project of intense research to make machines more trustworthy and interpretable (regardless of, or in parallel with a moratorium.) The premature super-investment in non-interpretable technologies is the core of our problems.
Quote Tweet
Absolutely, a Manhattan project on getting AI systems to align with human values would be a great thing. (As an aside, the letter wasn’t a blanket moratorium and it actually encouraged research to make machines more trustworthy and interpretable; I would not have signed… twitter.com/alexandrosM/st… Show more
30
653
Teller didn't "posit" this. They checked their math and realized it wouldn't happen. Bethe: I sat down and looked at the problem . . . I found that it was just incredibly unlikely . . . Teller was very quickly convinced and so was Oppenheimer. blogs.scientificamerican.com/cross-check/be
Quote Tweet
genuinely do appreciate the need for thoughtful concern when it comes to AI, but in 1942 edward teller posited a single atomic blast could ignite the earth’s atmosphere, trigger a runaway fusion reaction, and turn our world into a small new sun twitter.com/paulg/status/1…
Show this thread
9
195
Show this thread
Useful comment, in general. Don't ask calibrate of your own neurotic reaction. Ask for more facts.
Quote Tweet
People sometimes ask me "How freaked out should we be about AI?", and... it's such a weird question to me? You want me to tell you how to feel? How about I explain what's going on, and you can have whatever emotional reaction seems appropriate twitter.com/wolfpupy/statu…
Show this thread
Don't miss part 2 of our conversation with ! Liv joins us to discuss #GPT4, the AI arms race, and why this might just be the most important century. Catch the full conversation here: youtube.com/watch?v=PO-et7
Quote Tweet
Be sure to catch the latest episode of the FLI podcast! @Liv_Boeree joins @GusDocker to speak about Moloch - the personification of lose-lose situations - and what game theory can teach us about beauty filters, science, and AI. 🔗 youtube.com/watch?v=XZYMIT
16
125
本日、「日本維新の会」の国会議員の先生方による寄付として、20 台のトヨタ Hilux ピックアップ トラック及び大量の魚とパンの缶詰がウクライナに発送されました。🇺🇦国民への支援・支持につき、日本の与野党の共同立場を高く評価し、心より感謝いたします。
Image
Image
60
2,872
I was part of the red team for GPT-4 — tasked with getting GPT-4 to do harmful things so that OpenAI could fix it before release. I've been advocating for red teaming for years & it's incredibly important. But I'm also increasingly concerned that it is far from sufficient. 🧵⤵️
69
3,557
Show this thread
This is not human. But it's at least half way there.
Quote Tweet
Illustrative example 1: Bing improves its writing by just being told to read Vonnegut's advice on writing (and offers a surprisingly "personal" take on how it applied those tips) twitter.com/emollick/statu
Show this thread
The Abe shooting suspect built his own firearms in a one room apartment. The sound of the machinery was very loud and neighbors reported it to the building management company, which responded by...putting fliers into every mailbox asking residents to be quiet and considerate.
Image
27
1,799
If opinions on one subject can be predicted from opinions on another, might be an ideology. When people think for themself their conclusions aren't predictable. Japanese have ideologies, but they're not my familiar ones, and it's so often so refreshing...
Quote Tweet
One time back in the mid 2000s I asked a Japanese friend if she thought abortion should be legal. She said: "No." Then she thought about it for a second and said "...Actually, yes."
Show this thread
"Currencies, gold, and bitcoin [are all] just liquidity on my balance sheet which I want to use to invest in productive assets run by prudent individuals who are trustworthy and honorable." Great essay:
1
1