diff --git a/R/spi.R b/R/spi.R
index bbd7bc885..e5196c6da 100644
--- a/R/spi.R
+++ b/R/spi.R
@@ -18,8 +18,7 @@
 #' than the HDI, because, the _"HDI can be noisy (that is, have a high Monte Carlo error)"_
 #' (Liu et al. 2015). Furthermore, the HDI is sensitive to additional assumptions,
 #' in particular assumptions related to the different estimation methods, which
-#' can make the HDI less accurate or reliable (see also discussion
-#' [here](https://twitter.com/betanalpha/status/1479107186030624771)).
+#' can make the HDI less accurate or reliable.
 #'
 #' @references
 #' Liu, Y., Gelman, A., & Zheng, T. (2015). Simulation-efficient shortest probability intervals. Statistics and Computing, 25(4), 809–819. https://doi.org/10.1007/s11222-015-9563-8
diff --git a/man/spi.Rd b/man/spi.Rd
index d5f2adbc5..085be3a30 100644
--- a/man/spi.Rd
+++ b/man/spi.Rd
@@ -78,8 +78,7 @@ uncertainty of (posterior) distributions. The SPI is said to be more stable
 than the HDI, because, the \emph{"HDI can be noisy (that is, have a high Monte Carlo error)"}
 (Liu et al. 2015). Furthermore, the HDI is sensitive to additional assumptions,
 in particular assumptions related to the different estimation methods, which
-can make the HDI less accurate or reliable (see also discussion
-\href{https://twitter.com/betanalpha/status/1479107186030624771}{here}).
+can make the HDI less accurate or reliable.
 }
 \note{
 The code to compute the SPI was adapted from the \strong{SPIn} package,