Interestingly think I just came to the opposite conclusion after building CLIs + MCPs for code.deepline.com
Where MCPs fit in - short answer is enterprise auth for non-technical users.
CLIs (or APIs + Skills) are easier + faster to set up, UX is better for most use cases, but a generalized API with an easier auth UX (in some cases, usually the MCP Oauth flow is flaky too).
So feels like an imperfect solution, but once you start doing a ton of enterprise auth setups, MCP starts to make more sense.
Only things that have made a measurable impact for me - marketing is sales:
1. Mom Test - https://www.momtestbook.com/
2. Do more sales calls + customer interviews. Then rewatch the call and debug what you're doing, saying. What is landing and not landing. I miss the majority of the signal while I'm on the call - I don't actually get anything from it until I rewatch/see how people react to different things I say.
3. Rinse, iterate, and repeat. Your goal is to solve a customer problem. How you describe it, market it, etc. depends on how a customer describes the problem, where they spend their time, when and how they find solutions to their problem.
You're 100% right - but I do think there are more lower accuracy cases than I initially expected, *especially* if you assume a human-in-the-loop. Still 10x better than status quo.
Ex. Content generation + zero-shot classification/mapping are powerful, and with a human in the loop (somewhat) responsible for accuracy, they can move much faster.
I've been dealing with this a lot recently and seems to be a generational difference. Surprisingly close age wise with completely different responses.
Anecdotally, people ~40+ do take it personally as parent comment mentioned. It's a "your time is more important than mine" feeling by putting the onus on them. Around ~40% of the time, more senior people have ignored a booking link, and responded with times that work for them, requiring additional messages to confirm everything.
People under ~40 almost always use the link without issue, and usually send a confirmation email or message. People <30 don't respond to the message, just book and I assume we're good to go.
Now I use a "Let me know if there's a good time for you. Otherwise, here are 3 slots and my booking link" message.
Even had to make a Google Sheets app script to list my open 1.5+hr slots for the next 3 weeks to help with pre-populating that.
On the mentor note, even if you don't have a formal mentor relationship, most people are open to discussing their roles and responsibilities.
Depending on the firm, even more would be ecstatic to talk about a specific idea or side project to provide their input on marketing/strategy/finance. As long as you come from a place of genuine curiosity and not "tell me exactly what to do for free," almost everybody wants to help you.
1. If a BLM post is all it took to "lose a friend," I'd reconsider your definition.
2. Not everything is about you. Your friends explicitly were not thinking about you when they picked a side.
3. You won't be convinced, but someone else may benefit from a tool I've used to get past this. Would you hold the same beliefs if you were born black in an inner city? Would you want the status quo in terms of police strategies? To be scared for your life at every traffic stop? It's easy to defend a stance from a position of privilege (directly opposed to "my safety".)
You can do the same for your friends with different political beliefs - most stances are based on someone's life experiences or environment. There's likely a reason they believe something, and it has nothing to do with you. If you grew up in their situation, or friends, or life experiences, do you think you'd still hold your same beliefs? Most of the time the answer for me is no, and therefore the beliefs have nothing to do with the person's character.
You're right, caps lock is the real enemy. How could I have been so wrong? Let's ignore everything because of that. Bars should open, as long as everyone whispers we're fine.
I can tell you're a smart guy, but it reads like you're buying into the politicization of an apolitical problem.
The end condition is a therapeutic/vaccine, and since that is not available yet, the next option is a national testing and contract tracing plan. Every successful country has a good one, every country that's seeing mass casualties doesn't have one. There is no political aspect to this. The science is pretty clear here (like climate change clear, meaning there's always a few contrarians for the sake of being contrarian, everyone else agrees). Herd immunity is not feasible, and millions will die.
I'm ignoring politicians, every annoying tech bro (always dudes...) who led "Growth" for a SaaS startup and therefore a are an expert in "virality" and "k-factor". Ignore Elon, this isn't his zone.
Everyone is bitching about the stay-at-home. IF THERE IS A CONTACT TRACING PROGRAM, HEALTHY UNEXPOSED PEOPLE WILL BE ABLE TO LIVE THEIR LIVES. This point seems lost on the people protesting. There's an exit strategy that their leader (protesters are largely Republican) doesn't want to do because testing reduces his re-election chances (?!??!?!). However, they have a testing and contract tracing program FOR THE WHITE HOUSE RIGHT THIS SECOND.
I've spoken to American friends in Hong Kong and they're so happy they live in a functioning autocracy. They're going out normally and enjoying their lives. Getting dinner with friends, going to beach, going out in LKF. (With occasional issues like the Seoul club superspreader). The only reason we are not having fun right now is because of the lack of testing and tracing.
It's expensive, intrusive, and people who get sick will be pissed and not follow the rules - but I don't see another option on the table, and none of the experts I've read or listened to (Fauci, Yaneer's team below) have another solution. This is not a poltical problem.
Yaneer Bar-Yam has been doing great work on supporting countries figure out the best plan, as well as preparing societies for the inevitable fallout of all of this. You hedge for risk.
https://twitter.com/yaneerbaryam
In the playbook left by the previous admin, the 3rd (5th in the domestic part) question in every stage of pandemic response, after "How bad is the virus and how quickly is it spreading?" is "Does the government have tracing and testing set up?". When it was a credible threat (January if not earlier), according to the extremely clear and again, NONPARTISAN, document below, we should have started setting up a testing/tracing program.
https://assets.documentcloud.org/documents/6819268/Pandemic-...
Caveat is that 15% of our nation lacks empathy and is completely detached from reality, and think wearing a mask is communism, so we'll never get to full compliance. However, 90% in a testing/contact tracing environment is better than what we have now.
In summary - if you want to end stay-at-home, you should hope the Trump admin magically becomes competent. Blue states will try to do it, but because of interstate commerce and conservative governments who think they're invincible for some reason, it's 100x harder.
Otherwise we'll be like this for what feels like forever and ~200k disproportionately lower-income, older and minority people will die. If he's re-elected in November, I'd put significant money on 500k+ deaths in the next 4 years. It won't go away completely and anti-vaxxers will feel vindicated by his re-election.
What percent of these roles do you think will actually go to local talent vs. be off-shored? If tech companies can outsource, than who is benefitting?
Related - Have you personally hired an H1-B employee? The H1-B process is a pain for employers. There's a lot of costs delays, uncertainty, and legal fees to hiring H1-B.
I'd be careful with package tracking. While I don't know if that particular app resell your purchase data. That one is paid, so they may not need to right now, but they may change that in the future. I know many other similar (Slice, Unroll.me, etc.) products do. It's an easy, high ROI monetization strategy.
I also agree. One-on-one mentoring is fulfilling in a different way than project oriented volunteering. I recommend doing a longer term commitment as well. Most are 2-4 hours a week, ideally for at least a year.
I was at a large near-IPO startup, and it was super supportive of me hopping out for 3 hours every Wednesday for a year to volunteer at underresourced schools. I brought my classes in to meet people at my company and many kids became more interested in design & programming careers as a result.
I worked with build.org in NYC - it's entrepreneurship oriented programs. The students are wonderful. BUILD is large and has a strong presence in SF, DC & Boston as well.
In SF, I worked with Reach & Rise through the YMCA, which is similar to the Boys & Girls Club Big Brother/Big Sister program. I'd highly recommend that as well.
Look into affiliate marketing. If your niche has any businesses in it, chances are you could potentially get as much if not more on a revshare model. Slightly more upfront setup with partnerships, but a learning opportunity nonetheless.
Where MCPs fit in - short answer is enterprise auth for non-technical users.
CLIs (or APIs + Skills) are easier + faster to set up, UX is better for most use cases, but a generalized API with an easier auth UX (in some cases, usually the MCP Oauth flow is flaky too).
So feels like an imperfect solution, but once you start doing a ton of enterprise auth setups, MCP starts to make more sense.