Gooch & Housego is not currently an AI photonics investment, because thin-film lithium niobate plays no role in co-packaged optics today. All major CPO deployments use silicon photonics; TFLN only becomes relevant as a supplement at 1.6T or 3.2T, likely a 2028+ play, and whether G&H will be competitive remains to be seen. That pushes back on @PhotonBull's case for G&H as the best TFLN pick. It is a solid company with strong defense tailwinds, and AI relevance may come in late 2027 or 2028. x.com x.com





