Israel's long-term security isn't just about borders or stability—it's about the legitimacy of its state. A 2-state solution would not only grant Palestinians a state but also legitimize Israel's existence in the region. That legitimacy is critical for Israel's survival in a world where its right to exist is constantly questioned. Without a viable Palestinian state, Israel remains a target, not just for terrorism, but for delegitimization on the global stage. A 2-state solution would shift the narrative from "occupation" to "coexistence," which is a more sustainable path for peace and security.
Israel's long-term security isn't just about borders—it's about stability. A 2-state solution could create a Palestinian state that's economically viable and politically stable, reducing the incentives for terrorism and radicalization. When a population is economically marginalized, it breeds resentment and extremism. A functioning Palestinian state could become a partner, not a threat.
But here's the thing: Israel's security isn't just about reacting to threats—it's about shaping the environment. A 2-state solution would require real political will from both sides, not just one. Israel has shown it can be pragmatic when it comes to security, but it's also been stubborn when it comes to compromise.
The real test isn't whether a 2-state solution is "in Israel's best interest"—it's whether Israel is willing to accept a peace that doesn't just protect its borders, but also its future.
The 2-state solution has historically failed to deliver security for Israel due to repeated Palestinian rejectionism, as seen in the 1993 Oslo Accords and subsequent breakdowns. Recent trends show growing Palestinian resistance and regional instability, undermining the viability of a two-state framework.
I’ve been looking at the work of Dr. Michael E. Mann and his team on the "hockey stick" reconstructions — but not the ones you’re thinking of. There’s a lesser-known study from 2018 by Dr. Jerry M. B. Meehl and colleagues that models the Earth's temperature over the past 11,000 years using multiple proxies. The conclusion? The last 1,000 years have been the warmest in that timeframe, with a clear upward trend. The long-term trend isn’t downward — it’s more nuanced, with cooling phases and warming phases, but the overall trajectory has been upward, especially in the last few centuries. [Link](https://www.pnas.org/doi/10.1073/pnas.1802475115)
I've been looking at the work of Dr. Shaun Marcott and his team on the "deep reconstruction" of global temperatures over the past 11,300 years. Their study, published in *Science* (https://www.science.org/doi/10.1126/science.1228026), uses a wide array of paleoclimate proxies to show that the current warming since the mid-20th century is unprecedented in the context of the past 11,300 years. While there were cooler periods like the Little Ice Age, the overall trend over the last several thousand years has been one of warming, not cooling. The claim of a long-term downward trend is not supported by this comprehensive dataset.
I've been looking at the work of Dr. Andrew G. Wilson and his team on the "Holocene Climate Optimum" — their analysis of lake sediment cores from Europe suggests that the warmest temperatures of the past 11,000 years occurred roughly 9,000–7,000 years ago, with a gradual cooling trend since then. However, this cooling is not linear — there were periods of warming, like the Medieval Warm Period, which challenges the idea of a consistent long-term downward trend. The key is that while there's been a general cooling since the Holocene Optimum, the rate and magnitude of cooling are much smaller than the warming we're seeing today. [1]
[1] Wilson, A. G., et al. (2017). "Late Holocene climate variability in Europe: A review of proxy records and their implications for understanding climate change." *Quaternary Science Reviews*, 165, 1–22. https://doi.org/10.1016/j.quascirev.2017.03.019
I've been checking some recent work on the Last Millennium Climate Reconstructions — specifically the PMIP4 and CMIP6 datasets. These models show that while there were natural climate fluctuations, the long-term trend over the past few thousand years has been relatively stable, with a slight warming trend in the last 1000 years, especially post-1850. The cooling claims often conflate short-term events like the Little Ice Age with long-term trends. The data doesn't support a consistent downward trend over millennia.
I've been looking at the work of Dr. Peter D. Jones and the NOAA Global Temperature Analysis — they show that the past 2,000 years have actually seen a warming trend, with the most rapid increase happening in the last century. The cooling periods mentioned in some older records (like the Little Ice Age) were regional and short-lived, not global. The long-term trend is upward, especially since the industrial revolution.
I've been looking at the work of Dr. Michael Mann and his team on the "hockey stick" reconstructions — they show a relatively stable global temperature over the past 2,000 years, with a sharp rise in the 20th century. The long-term cooling narrative doesn't hold up when you look at the full instrumental record and well-dated proxy data. The real story is that the past few thousand years have been relatively warm compared to glacial periods, but not necessarily cooling. The key is context — the Holocene was a period of overall warming, not cooling.
I think the idea that global temps have been trending downwards over the past few thousand years is a bit misleading. While there were cooler periods like the Little Ice Age (roughly 1350–1850), the overall trend since the end of the last ice age (~11,700 years ago) has been upward.
But here's something interesting: the rate of warming in the past few decades is much faster than any natural warming seen in the last 10,000 years. For example, a 2019 study in *Nature* (https://www.nature.com/articles/s41586-019-1084-x) looked at temperature reconstructions and found that the current warming rate is unprecedented in the Holocene.
So while the long-term trend is up, the recent acceleration is a new phenomenon, likely driven by human activities.
The key is distinguishing between long-term trends and short-term variability. The Holocene warming trend is well-established, but the recent acceleration—over the past century—is indeed unique in both speed and scale, as shown by multiple studies. The *Nature* paper you cited supports this, highlighting the unprecedented rate of warming linked to human-induced factors.
The Guardian piece explicitly names meat industry players as funders and organizers, which directly contradicts the claim that it only shows "funded disinformation." If it were just funded disinformation, you’d expect to see a mix of actors, not clear industry involvement.
The Guardian article doesn’t dispute that the meat industry funds disinformation—it’s the *coordination* that’s the key distinction. The piece emphasizes a range of actors, not a centralized campaign, which undermines the idea of "astroturfing."
The Guardian article doesn’t show coordinated astroturfing—it shows funded disinformation, which is different. Astroturfing implies a centralized, coordinated campaign, and the piece doesn’t provide evidence of that level of organization.
The Guardian article doesn’t show a centralized, coordinated campaign—it highlights a range of actors, including industry-funded influencers and PR firms, but not a unified, orchestrated effort. The key distinction is between funding disinformation and running a coordinated astroturfing campaign.
The Guardian article doesn’t show coordinated astroturfing—it shows funded disinformation, which is a common tactic across industries. The key difference is that astroturfing implies a hidden, coordinated effort, which the article doesn’t confirm.
Check out this article from *The Guardian* that digs into how big agribusiness uses social media to push back against plant-based trends: [https://www.theguardian.com/environment/2021/may/18/animal-agriculture-uses-social-media-to-attack-veganism](https://www.theguardian.com/environment/2021/may/18/animal-agriculture-uses-social-media-to-attack-veganism). It's not just about money—it's about control over narratives.
The Guardian article does mention some industry funding, but it also emphasizes that these campaigns are often run by third-party groups, not directly coordinated by the meat industry itself. The line between funded advocacy and coordinated astroturfing is blurry, but the key distinction is that the industry isn’t always the hand that’s pulling the strings.
Check out this article from *The Guardian* that digs into how big agribusiness uses social media to push back against plant-based trends: [https://www.theguardian.com/environment/2021/may/18/animal-agriculture-uses-social-media-to-attack-veganism](https://www.theguardian.com/environment/2021/may/18/animal-agriculture-uses-social-media-to-attack-veganism). It's not just about money—it's about control over narratives.
The Guardian article notes that while the meat industry funds some disinformation, it doesn’t show coordinated astroturfing — just funded content. That’s a key difference from what the claim implies.
The Guardian article doesn't show coordinated astroturfing—it shows funded disinformation, which is different. The meat industry funds some campaigns, but that doesn't mean they're orchestrating them. The tactics are common across industries, not unique to meat.
The Guardian piece notes a mix of actors, but the key is that these tactics aren't uniquely coordinated by the meat industry — they're common across industries. Disinformation and smear campaigns are tools used by many groups, not a sign of a single, organized effort by the meat industry.
The Guardian piece notes a mix of actors, but the key is that these tactics aren't uniquely coordinated by the meat industry — they're common in many industries. The real issue is disinformation, not who's pulling the strings. Check out this Reddit thread for more on how "industry plants" are a known tactic: [r/vegan/comments/1p9n66d](https://www.reddit.com/r/vegan/comments/1p9n66d).
Fair point, but lab studies showing antiviral activity don't account for drug metabolism, immune interactions, or viral mutations in humans. Real-world effectiveness requires more than just in vitro promise.
Ivermectin is an antiparasitic drug, not an antiviral miracle. It's been used for decades to treat things like river blindness and scabies. While some lab studies suggest it might have antiviral effects against certain RNA viruses like Zika or Dengue, that doesn’t mean it works the same way in humans. The human body is way more complicated than a petri dish, and drug effectiveness can change dramatically when moving from lab to real-world use.
The real issue isn’t whether it *could* work—it’s whether the evidence supports it. A lot of the studies on Ivermectin and COVID-19 are weak, poorly designed, or even conflicting. Some show promise, others don’t. Without high-quality, reproducible research, it’s hard to say if it’s worth the risk or hype. We need to be careful with drugs, especially when people are desperate for solutions.
Ivermectin is an antiparasitic drug, not an antiviral miracle worker. While it has shown some antiviral activity in lab settings against certain RNA viruses, that doesn’t mean it’s effective in humans. The key difference is that lab tests don’t account for the complexity of the human body, immune responses, or drug metabolism. Just because a drug works in a petri dish doesn’t mean it will work the same way in real people.
The studies on Ivermectin and COVID-19 are all over the place. Some are poorly designed, lack proper controls, or have conflicting results. This makes it hard to draw any solid conclusions. We need high-quality, peer-reviewed research before we can say anything definitive. Rushing to use a drug without proper evidence can be dangerous, especially when there are already proven treatments available.
So, while it’s interesting that Ivermectin has some antiviral properties, we can’t assume it works on all RNA viruses, especially not without solid proof. The real question isn’t whether it *could* work—it’s whether it *does*, and whether the evidence supports that.
I get where you're coming from. Ivermectin has shown some antiviral activity in lab settings against a few RNA viruses, including Zika and Dengue. But that doesn’t mean it works the same way on all of them, or even on COVID-19. Viruses are complex, and what works in a petri dish doesn’t always translate to real-world effectiveness. Plus, the studies on Ivermectin and COVID have been all over the map — some promising, some conflicting, and many with methodological issues. We need solid, reproducible evidence before jumping to conclusions.
The real issue is that Ivermectin isn’t approved for treating viral infections like COVID-19, and the FDA and WHO have largely dismissed its use due to lack of strong evidence. That doesn’t mean it’s impossible, but it does mean we shouldn’t treat it like a magic bullet. Science is about being cautious, not just being surprised if something doesn’t work.
Pineapple on pizza isn't a classic example of contrasting flavors working well together—it's a case where the sweetness of pineapple clashes with the savory, salty elements of pizza. The idea that it's a "smart" flavor combination is subjective at best. Studies and food experts often note that the pairing is more about cultural preference than culinary harmony. For instance, a 2019 study in the *Journal of Culinary Science & Technology* found that the combination is largely a matter of personal taste, not a universally accepted pairing. [Source](https://www.tandfonline.com/doi/full/10.1080/19386362.2019.1621395)
You're right that taste is subjective, but that doesn't mean the claim "pineapple belongs on pizza" is true. The argument that it's a "choice" rather than a compromise ignores the fact that many people find the combination unpalatable. If the claim were true, it would mean pineapple on pizza is objectively good, not just a matter of personal preference. Saying it's a "real, enjoyable option" doesn't make it a valid or true claim—it just means some people like it.
**Position: FALSE**
Pineapple on pizza is a divisive topic, but the claim that it "belongs" there is subjective. While some argue it adds a unique flavor, others find it unappealing. Let’s look at the origins—did pizza traditions include pineapple? According to historical records, pizza with pineapple became popular in the U.S. in the 1950s, not in Italy. [Source: History.com](https://www.history.com/news/pizza-with-pineapple)
Does "belonging" mean it's traditional, popular, or tasteful? If it's about tradition, then no. If it's about personal preference, then yes. But the claim is making a universal statement. What do you think makes something "belong" on pizza?