Assessment: The prediction matches what actually happened between mid‑2021 and roughly mid‑2023.

1. "A lot more emerging variants" of SARS‑CoV‑2

   • After June 2021, the world saw the emergence of multiple new variants and subvariants. Most notably, the Omicron variant was identified in November 2021 and rapidly became globally dominant, followed by a large family of Omicron sublineages (BA.1, BA.2, BA.3, BA.4, BA.5, BQ.1/BQ.1.1, XBB.1.5, etc.). (en.wikipedia.org)
   • Through 2022–2023, additional Omicron subvariants such as XBB lineages and BA.2.86/JN.1 continued to emerge and spread, illustrating exactly the pattern of “a lot more…emerging variants” over the “next couple of years” after mid‑2021. (en.wikipedia.org)
   • Even beyond that two‑year window, new subvariants (e.g., XFG in 2025) have continued to appear, reinforcing that the virus kept evolving into distinct global lineages. (indiatimes.com)
     This clearly fulfills the first part of the prediction.

2. "Increasing number of risks from potentially engineered biological agents"
   While there has not been a confirmed large‑scale engineered bioweapon attack in that specific period, the risk environment due to engineered or engineerable agents has measurably intensified, which is what the prediction referred to:

   • The Biological Weapons Convention community and expert analyses explicitly note that rapid advances in synthetic biology, gene editing, and enabling technologies are eroding the technological barriers to acquiring or genetically enhancing dangerous pathogens for hostile use. (en.wikipedia.org)
   • Unclassified U.S. intelligence threat assessments (summarized in legal and media analyses) warn that rapid advances in dual‑use technologies—bioinformatics, synthetic biology, genomic editing, etc.—could enable development of novel biological threats, i.e., new or engineered agents. (americanbar.org)
   • Academic and policy work over 2022–2023 highlights that AI, large language models, and biological design tools can lower barriers to designing or optimizing dangerous biological agents, explicitly framing this as an increasing biosecurity risk from engineered biology. (arxiv.org)
     Collectively, these official and expert sources converge on the view that the risk from engineered or engineerable biological agents has been rising since around the time of the prediction, consistent with Friedberg’s claim about “increasing…risks.”

Conclusion:
Both components of the prediction came true in the relevant timeframe: many more SARS‑CoV‑2 variants did emerge globally, and expert and official assessments confirm a growing risk landscape around engineered biological agents. Hence the prediction is best scored as right.