Confirming previous forecast. No new information to indicate otherwise.
0.63875
Relative Brier Score
76
Forecasts
86
Upvotes
Forecasting Calendar
Past Week | Past Month | Past Year | This Season | All Time | |
---|---|---|---|---|---|
Forecasts | 0 | 4 | 87 | 25 | 342 |
Comments | 0 | 0 | 14 | 1 | 62 |
Questions Forecasted | 0 | 4 | 16 | 9 | 67 |
Upvotes on Comments By This User | 0 | 2 | 88 | 13 | 173 |
Definitions |
Lowering due to passage of time. No buzz indicating any progress on technical feasibility of labelling AI-generated text, and with the question deadline looming large, lowering by several percentage point to reflect that reality.
FYI, interesting read:
Drawing from information theory, we argue that as machine-generated text approximates human-like quality, the sample size needed for detection increases. We establish precise sample complexity bounds for detecting AI-generated text, laying groundwork for future research aimed at developing advanced, multi-sample detectors. Our empirical evaluations across multiple datasets (Xsum, Squad, IMDb, and Kaggle FakeNews) confirm the viability of enhanced detection methods. We test various state-of-the-art text generators, including GPT-2, GPT-3.5-Turbo, Llama, Llama-2-13B-Chat-HF, and Llama-2-70B-Chat-HF, against detectors, including oBERTa-Large/Base-Detector, GPTZero. Our findings align with OpenAI's empirical data related to sequence length, marking the first theoretical substantiation for these observations.
Active Forecaster
Updating with some new movement after the 7.4 magnitude earthquake.
UK, Canada and Australia made note of the earthquake but did not issue a Level 3 above or equivalent advisory.
Updated:04 April 2024
Latest update :A 7.4 magnitude earthquake occurred off the coast of Taiwan on 3 April. There's damage to buildings and infrastructure, with disruptions to local transport. Avoid affected areas and follow the advice of local authorities. (See 'Safety').
We advise: Exercise normal safety precautions in Taiwan.
See also:
Started this question with a better than even chance, lowering to far less than even. #Diplomacyishard
"Two American officials agreed that the prospects are not promising of Israel and Hamas agreeing to the temporary truce by the start of the Muslim holy month early next week. " (CNN)
I am intrigued by Biden saying that the impetus is on Hamas. According to CNN, Hamas has demanded that "...in the first phase the Israeli military would pull back from Gazan cities and in a second withdraw from the enclave altogether, according to the diplomat who said the IDF is refusing to agree to those points." After months of conflict, continued violation of IDF's own promises to mitigate harm to civilians, it is hard to see an end-state that can fully accommodate the IDF's stance (and maybe that is the intention).
The only practical means to a ceasefire would be for the US to use more from its toolbox, not just airdropping supplies to Gaza (reducing/cutting arms assistance for example). And the political feasibility of that is bleak.
Iran wants a ceasefire, the US is currently negotiating one (albeit with limited success). An (attributed) Iranian airstrike will be counterproductive.
Will track the ceasefire negotiations, and revise if those fail.
Also re-upping previous rationale, with the addendum that according to WaPo, Tehran has reportedly urged the groups it is backing to exercise restraint.
"Many see the Houthis as Iranian proxies. But this is only part of the story, as the Houthis are also using their ties with Iran to advance their own interests and have been keen to expand relations. They have been under attack by a military force far superior in conventional terms, and only Iran has been willing and able to provide external support. But this does not make them pawns of Iran. There is no evidence, as some Iran hawks allege, that the Houthis take orders from Tehran or would have adopted significantly different policies absent Iranian support."
Iran, and most of the Middle East, really just wants to region to stabilize, and for a feasible two-state solution to come out of this conflict. And the United States' role in bringing this conflict to a close is crucial (and grounded in historical precedents), because Israel really has no other friends in the region. Unless there is a fundamental change in this calculus, either in DC or in Tehran, a deliberate attack is highly unlikely.
No updates since the news in January. Lowering "Yes" by a couple of percentage points.
It's the year of AI chips, evidenced by Nvidia's rapid growth in value, inching closer to Apple, which remains the world's most valuable company by a relatively slim margin.
"Last month, the chip maker reported a 769 percent increase in fourth-quarter profit and a 580 percent jump in full-year profit compared to the previous year."
TSMC and Nvidia are the leaders in AI chips, and according to Commerce Secretary Gina Raimondo, the AI boom has generated massive demand for chips.
With all this in mind, remaining highly optimistic for 2024.
https://www.nasdaq.com/articles/nvidia-might-soon-surpass-apple-as-worlds-second-most-valuable-company
Lowering with passage of time. No news since the flurry of articles in January indicating that "Microsoft has no plans to close down its AI lab in China".
Also, Microsoft's Beijing office, which is where the research lab is housed, is actively hiring:
https://careers.microsoft.com/v2/global/en/locations/beijing.html