A 95% confidence interval for a mean is a range built so that, over many samples, 95 percent of such ranges would contain the true mean. It is the estimate plus and minus a margin:
mean ± t* × SE, where SE = SD / sqrt(n) and t* comes from the t distribution with n - 1 degrees of freedom.
Use t, not the normal's 1.96, because the SD is itself estimated. With 50 loaves the difference is small; with 8 test bakes it is large. In SciPy: stats.t.interval(0.95, n - 1, loc=mean, scale=se).
A wider level (99%) gives a wider interval; a larger sample gives a narrower one. The interval is about the average, not about single loaves: most loaves fall well outside it.
If a promised value, such as 800 g on the label, lies inside the interval, the data can't rule it out.
A 95% t interval for the mean loaf weight.