Roadmap feedback

Hello, our PreGPT is currently in beta with some community and team members. We have an internal roadmap on this matter, we hope to share more information about it in the coming weeks.

will GPT3 still be available along side the salad AI?

Yes, The current gpt chat version that is inside Presearch will still be there, Pregpt will have a different access (domain) that you can enter with your Presearch username and password.

Will it be available inside Presearch as well like an option to add for search results?

I would like to have the new Pre IA providing quality results on my searches right next to or above the GPT3 results. Then if I want to chat with it I can go either to the separate login or just have a slide over window to interact with it.

Having this AI on searches is what could improve the search experience for users who choose to signup for the service. Since the service is tied to the Presearch user you should be able to easily continue the conversation separately if needed.

Below is what I think would make this Salad AI a great addition to Presearch user experience. If it is a completely separate domain and website that does nothing to improve SEARCH, there needs to be direct integration with search results!


1 Like

Thanks for the contribution, what has been said is that it will be a different domain or site than presearch.com but you will enter with your Presearch password and username

I wonder how many people that took the survey thought it was going to be a completely separate service and domain than presearch.com?

For the sake of salad and presearch I hope there is an API or direct access on the presearch platform with search results.

Even if direct 2-way connected follow-on or deep dive query chat is only available on the separate domain that would be completely fine as long as there is an option to get quality AI results for your logged in presearch queries, just like the current GPT3. I also wouldn’t care if the search query and salad results were not connected or carried over to the other separate domain. In fact it would probably be cleaner and more privacy focused if the search query results were not connected.
If you want to do back and forth with the salad chat then you just login on the other domain and go at it.

I just strongly believe every effort the team makes should be hyper-focused on improving the presearch platform and user experience. Additional quality ai search results would do that.

I think that no one who filled out the survey thought about that, it was not the central topic, It is still in beta phase and we do not know if in the future it will be linked in some way with presearch.com, Taking into account that chatgpt is already within the search engine

I view ai results as a substantial improvement over the legacy search experience and all reputable search providers will end up migrating towards having the option for ai results. The question is which ai will they use, how helpful are those ai generated results and if any will begin to offer a meta ai experience for improved search results. People will begin using search providers based on the quality of their ai results assuming all else on the traditional side of results are somewhat equal. I would say 95% of presearch traditional results are as good or better than Google. Assuming presearch will eventually close that gap on the other 5% things like maps they will be fine. (Maps I personally view as a lower priority than developing quality ai or meta ai search integrations. I say this because maps are a specific and tiny part of overall search queries whereas ai would improve all those other 95% queries) people will make their decisions to use a provider or not based on the quality and reliability of the ai results to get their immediate answers to queries and accomplish simple tasks.

GPT3 is a laggard literally using data from years ago. It’s still helpful though but compared to others it’s less helpful. Google, Bing and others are building in bias to their ai and although great for some queries are woefully inadequate on others. This presents an opportunity to offer better. You do that by not picking winners and losers but instead attempt to onboard all ai or as much as possible to create a meta ai for search.

What I mean is access to multiple AI providers for search query results in addition to the traditional result links. Salad would be a great first step in addition to GPT3. Then add Grork X’s ai once an API is available, GPT4, llama, etc. procure a wholesale usage pricing model then resell to your users at a daily, weekly, or monthly retail rate. Use the decentralized nodes to access the APIs returning results to users from a plethora of AI services all with anonymity and privacy and from a single innovative presearch platform. If you are subscribed to 2 or more AI services you would prioritize the order in which they show up on the results page followed by traditional linked results.

If the first ai does not provide a quality answer maybe another subscription would or it might even provide a counter bias to first. There will be thousands of ai services some open sourced others not, many will begin to utilize proprietary datasets which is why data is the new oil, others will focus on niche applications or industries and provide much higher quality than any general ai will ever be able to accomplish.

So what I am saying is this concept will eventually be developed but the first search provider that understands this future and is willing to build a platform that supports all ai will become the de facto leader. I believe that vision is worth developing I believe that would bring millions of users and hundreds or thousands of ai services to the platform only increasing the stickiness and brand awareness for the platform. I hope that platform is presearch. I certainly believe the decentralized node infrastructure lends itself to the concept.

I talked to Tim about this at length months ago and urged him to consider making presearch the innovative trail blazer. If you want people to notice the platform it needs to be changing the game not playing by other people’s legacy rules (if we play by their rules we will lose). If we innovate and improve people’s search experience they will notice and share that with others. Eventually the other providers will have to adapt to your game or continue to lose market share.

1 Like

Hello XRPRE, I really appreciate all the suggestions you make here and on Discor and about PreGPT I would like to give you a more detailed explanation about it, we are currently in beta, I will see how I can let you know.

1 Like

I think AI chat is great but there are already tons of competition and since search is what PRESEARCH is all about there needs to be a focus on leveraging AI services to improve SEARCH. AI can and does improve search so the simple add to the roadmap is the integration of AI services with Presearch query results.

When you signup for AI services (of which there should be many) through Presearch you get the added benefit of having the option to toggle on ai results for your regular Presearch queries. This makes every ai service more valuable when they offer through Presearch. Presearch just secures wholesale per user pricing to various ai services and charges Presearch users a retail rate. Then they get the full ai service PLUS they get the option on their Presearch queries to see their subscribed ai results as well as traditional linked presearch results all on one platform. There are many ways to implement this but most ai services have an API that could be accessed via the Presearch node network at least when using for search results. This provides an additional layer of security and privacy for search queries. The user would still always have the benefit of opening the ai full service in a separate window so they are not losing out on functionality they are only gaining value by subscribing through Presearch.

//

When you design your paywall for PreGPT:

-I would strongly consider a daily 24 hours (ex $1.5), weekly 7 days (ex $3), monthly 30 days (ex $7), and annual 365 days payment access options. This creates more accessibility in terms of pricing and will increase sales while spreading load usage of the service. This is obviously different and in addition to any planned access tiers. At the end of the day the longer duration subscriptions are more money but proportionately less expensive. More people will try and test and use both regularly and periodically.

-I recommend adding PRE payments via 3 options on the PRE payments page:

  1. Presearch main user wallet (probably will be the most common) this would spend tokens from your main Presearch user wallet.

  2. Presearch usage rewards. I think this adds a great value to usage rewards making them usable for certain consumptive uses on the Presearch platform like ai access. This is a common complaint from the community that “you give rewards through searches and referrals that are not usable and have zero value” unless you reach 1k or more tokens. This would begin to counter the argument that usage rewards are a farce, at least for consumptive uses, it would also reduce future PRE obligations to Presearch (of which I am sure there are millions of tokens in future obligations). Many users that are well short of 1k would see this as an opportunity to use what they perceive to be otherwise unusable tokens and get some value out of them. This will allow them to have less of a chip on their shoulder, more respect for presearch, and once they experience AI they may be new buyers of PRE to stake the 1k for GPT3 or start search staking or pay monthly for PreGPT. This also simultaneously returns PRE to the treasury. Why wouldn’t you offer the use of these usage reward tokens for consumptive uses on the platform, this is almost like a store credit?

  3. Payment from self-custody, this is less important for now but will be required in the future and might as well be considered for development as an option now to show the forward thinking and future when things such as Cosmos and other multi-chain options begin to come online and more people are using self-custody wallets.

//

Add option for users to pay for other users: This only increases potential sales.

-I would recommend adding an option on the PRE token payment page to both stake (for old GPT3 only) and/or pay (for new PreGPT) for another user’s access. Adding a promo code field might also be beneficial here and provide great value to Presearch where users could go to submit future promo or 1-off access codes.

The primary reason for this is because it creates increased use, additional onboarding/searches, and of course additional token locks and income. WHY? There are many potential users that either don’t understand or don’t want to utilize crypto, they also may not understand AI either. This would allow current and future Presearch users and long-time community members to expose their friends and family to both GPT3 or PreGPT by inputing their user name then either staking 1k PRE for GPT3 and/ or paying the PRE fee for PreGPT.

How this would work:
Once the username is typed into the field it could be your username (primarily for promo/access codes if this option is added, or another’s username. (Pay it forward, stake, pay, or input promo/access code for a user to get access to Presearch AI GPT3 or PreGPT or for promo/access code submissions) You then hit a verify button that confirms the username or email inputted as a valid user (if valid the username turns green if not it turns red and warns this is not a valid user) then you select either stake 1k, and/or pay, and/or promo/access code then submit. It gives you all the details stake and Pre Costs and where each will be paid from. Then you confirm the transaction.

For stakes on other users your PRE is still your PRE and is just being held in a new stake category for other users. You can even have a page that shows your other user stakes and the username you are staking for. If you ever need or want your tokens back simply reduce your stake to 0 and your tokens will be returned for your main Presearch wallet and that user will lose their ai GPT3 results. Fully in your control and fully in your hands to be able to benefit other users.

This gives community members more tools to influence onboarding people to Presearch. Family, friends, or anyone you think might be interested this provides you a more compelling reason to sign them up, download the app and get them using Presearch so that they can see the benefits of search with AI. Or experience unbiased AI chat etc. It’s already hard enough to change peoples habits, this provides the community a simple incentive tool at no cost to presearch other than initial dev time.

This promo/acces code option provides Presearch an avenue to reward certain behaviors or as part of competitions or periodic promotions. “Sign up now and get 25 PRE plus GPT3 or PreGPT access for 1 day, 1 week, 1 month, etc.” Then if it’s a new signup from that promotion an email is sent with the 1-time use promo code or it’s a time-based promo for all recently registered users. Presearch needs to give itself value-add options to influence people to try Presearch. If its a 1-time use code that someone wins and they already have a subscription this code can still provide value and be used on another user.

All of this only increases sales, increases token locks, and increases value of the presearch platform.

Roadmap feedback as requested:

PreGPT AI Chatbot Incubation Launch Phase
I reviewed the description and overall seems on-track. However, 2x key points to consider:

  1. I don’t see search integration. This is a huge gap and shortfall if not planned to be implemented. I suggest creating a linkage just like GPT4 where if subscribed you can turn on or off various AI search responses. If you want multiple AI responses turned on in the initial phase maybe only GPT4 and PreGPT you would order preference them to select the order you want their results shown. You can still have it default 1-2 lines shown with drop downs to see more of each. This will only increase user searches and provide better search results. If GPT4 doesn’t provide a quality answer but PreGPT does great you don’t even have to view any traditional search results but viewing both AI also gives you a better idea if the answer is accurate or not multi sources is usually better than single source. I think the same will be the case for AI. Getting this done also provides more value to the user because they can use the typical chat feature and/or use with their regular Presearch search queries. Developing this feature will then allow you to add other AI into the mix and become the ONLY SEARCH PROVIDER offering this decentralized service/capability of integrating multiple AI into search.
    In order to not get chat artifacts meaning the chat tries to answer the next search query using previous information based on the last query which could taint the accuracy and relevance of the result, you would need to ensure all search queries are treated as a new chat query. If you need the capability to dive deeper and point the chat in the direction desired or build on a previous answer then the user can just switch over to the chat feature. In the future I think the most seamless user experience would be a side slide bar off to the right that you could open or collapse for this utility.

  2. The other thing that needs to be done right with the chat service is the PAYWALL. I saw a preview of the stripe paywall and option for promos that is perfect and how it should be with CC or Apple pay options that will immediately charge the customer and immediately provide income to Presearch. The same type of paywall needs to be created for Paying in PRE. If you selected PRE as payment option it should pull a quote maybe every 15-30 sec based on the price of PRE and the cost of the service in fiat. Then show the exact amount of PRE you will be paying. Once you hit next it takes you to the confirmation page showing the service you are purchasing, any declaimers, your total available PRE balance, the total PRE cost, and total PRE balance after payment confirmation. Hitting the confirm button will draft the PRE and sweep it into the Treasury giving Presearch immediate PRE income and the user access to the product/service.
    Keyword staking Plugin
    This is great and much needed for advertisers to better track and manage. However, without significant revamp of Keywords this becomes wasted development effort. Keywords MUST BE REVAMPED. There is too much potential income and opportunity for ALL to continue to ignore this.
    • I STRONGLY URGE THE TEAM TO ADD KEYWORD REVAMP TO THE ROADMAP. I have already laid out what I think would easily 10x your current keyword stakes and income just build it and implement it. This would only take up a single text ad slot but provide so much value and still give you the option to offer other similarly placed or different ad types in the future. It’s a 1-time build and a lifetime of utility, USP, income, and TVL.
    Presearch Maps and Translate
    I assume the intent is just trying to close capability gaps with competitors to have a true 1-stop search option. This is good but probably shouldn’t spend too much effort, money, and time on it yet. Find an easy to integrate cheap API to provide a MVP in this niche area and focus efforts elsewhere like keywords and improving search with AI etc. The core product is SEARCH be innovative and focused on this. Other things like maps etc are extras and nice to haves and in most cases if used are often used through an app not the search platform itself. We have to have our priorities in order especially with limited funding and dev capacity, SEARCH IS IT. We need to get back to what search was meant to be or we fail. If this is going to be too costly or time consuming or a dev suck drop it until other higher priorities are completed.
    Community search engines – white labeled
    I am super excited about the prospect of community search engines. What it could do for adoption of the platform, what it could do for personalization of ads and bringing in new advertisers, and eventually what it could do for improving search results once storage and indexing are available and the community is able to curate results. I have laid out a vision for this in the past but this should be discussed as a community how best to implement. I am still in favor of an NFT representing community ownership. This NFT should in most cases require a stake and have an upfront consumptive cost. The NFT held in your wallet could then be the key to access and develop your community and gives you the potential benefits of community ownership. It could be transferred to whitelisted addresses in the future to record a new ownership for a transfer fee creating a community ownership marketplace. People could buy or sell communities in the future or mint new ones. If the minting fee (stake + consumptive costs) of new communities increases over time with demand it could make all communities more valuable for potential resale. Successful communities with low grandfathered stake and consumptive fee paid would maintain the fruits of a vibrant Presearch Community. Others (speculators, or unsuccessful community managers) may choose to hold and sell their community for a profit in the future. Community names should be able to be renamed to adjust to the culture of the community but all community names must be unique meaning if another community already owns the name no one else will be able to use the same community name.
    There should always be a benefit for all parties, meaning there should be a benefit to the user, to the community owner, to Presearch, to advertisers, and even to the broader presearch ecosystem. Lots to discuss and to be excited about with this roadmap item.
    Deeper Web3 Integration – No issues with description. Important for transparency and to maintain control of crypto stakes, once implemented would allow stake to be made from self-custody instead of centralized with withdraw limitations etc. Might eventually need to build-in an unlock cooldown period to prevent mass unstakes and sales around short term price fluctuations. This would also give the community and team some transparency as to total volume that might become future potential issues.
    Presearch mobile App – Love the iOS app and is probably ready for an update with latest brave tech. The mobile app is a great onboarding tool. I have had great success with making this the default browser for the phone which ensures all browser usage is then routing searches through presearch. Its based on the brave backbone so has the added benefit of striping all youtube and other ads keeping it fast and exceptionally useful. Whatever updates you do, Do NOT change this video ad blocker feature!
    Tradingview API – Would this replace the community search feature for showing coingecko price results on crypto tickers? If so great because that community search feature requires multiple attempts to trigger the price mechanism. I have mentioned this issue in the past several times and it still has issues. Sometimes I have to input the ticker 3-4 times before it triggers this is wasted/inflated searches from a true search number and advertising perspective.
    Ongoing Platform Optimization - Assume this is a placeholder description does not provide much context.
    Growth Hack Sprint – Ok. We definitely need growth but we also need to ensure core USPs are in place so we retain the new users and they become excited to share with others. We also need to ensure Node reward model is addressed so that there is not an immediate substantial growth of un-needed nodes. If the node reward model is addressed then bring it on, new nodes that are more of the same should not impact rewards of quality longtime node runners and both new and OG node runners can adjust accordingly. This change to node rewards should ensure decentralization and quality is upheld, but still provide a staked capacity component just to a lesser extent. Also think targeting warm markets that already distrust Microsoft, Google, etc is best bang for the buck at this time.
    Community / Influencer Content Campaigning – same advice as growth hack sprint.
    New Exchange listings – Great much needed but the reality is that it is much needed because it is currently only ERC-20 compatible the worst L1 aside from Solana in terms of usability. I think there will always be a need for CEX on and off ramps but DEXs and AMMs on superior networks will likely increase in usage at a similar or faster adoption rate than CEXs (in fact I think CEXs will eventually tie into and utilize the increased depth of liquidity available on the AMMs). Base is a good first step, but COSMOs, and others like XRPL, FLARE, XLM, ALGO, etc. will open up more DEX and AMM options. Especially focus on regulatory compliant or legal clarity. Additionally, if the team focuses on increasing utilities for the PRE token demand and usage volume will increase. Especially, if offering fiat payments that require or give the option for PRE use this means Fiat is used for a good or service but Presearch taps into liquidity on CEC or DEX or AMM to purchase behind the scenes on behalf of user. Once you get volume pumping the token will be listed everywhere without having to pay for listings.

1 Like

Feedback continued:

Support for X.com Integrations - Description says placeholder for creator economy – assume this will get push right to another Quarter. I also think this is pending X.com money transmitter licensing for the last 15 US states and how they plan to integrate crypto payments, tipping etc on the platform.
1st Annual Community Meetup – Please don’t waste any limited funds on this. Spend funds on development of search and improving the platform. I am not saying Community meetups are not important I just think they should be no host socials. Plan to meet at a location at a specified time and everyone pay their own drinks food etc.
DAO Formed – Will be interesting to see when this becomes available and on what chain? I would recommend not launching on ETH unless the code can be easily ported to another chain when available. --It is not clear what the vision is for the DAO and how it will function (will it be simple PRE voting? would this require WrappedPRE? Will users without PRE get a vote or any vote power? Will votes all be transparent and on chain? Will there be multi-chamber voting like a user/PRE general chamber and a trusted DAO board with team and trusted community members? How do we prevent proposals that are not feasible for implementation?) I think this needs community input and discussion prior to development and rollout unfortunately have not heard any details yet.
Collaborative Gamification – lots of ideas in the description looks like some good ways to keep people active and with stripe addition of promos the rewards don’t all have to be inflationary PRE give aways. – I think best thing for adoption is continued development. Once you get enough great USPs and utilities that are innovative and change the search and advertising game for the better towards decentralization not centralization, these are the things that will bring people over and keep them. If you market to cold markets and don’t have USPs they care about they will not be converted and marketing funds will have been wasted. If the USPs are developed and/or enhanced along with increased PRE utilities this are the areas gamification should focus on.
Cosmos PreChain Phase 1 – Great lets get this deployed already. ETH sucks and can’t scale, Cosmos seems like a decent chain to launch on and would give access to Cosmos ecosystem of chains as well as Axelar chains (which will include XRPL soon). At the end of the day just need to continue to develop and get PRE on other chains. No chain will be the end all be all as they are have their own unique selling points so the more chains the better and opens different utilities. Example: XDC is EVM compatible and works just like ETH but is faster and cheaper great for smart contracts and trade finance. XLM has smart contracts now too and AMM at the protocol level. FLARE is becoming more interesting now due to trusted decentralized bridge with collateral backing, ability to cross chain to others using layercake, their decentralized data feeds bringing different data sources on chain in a decentralized way, on-chain AI needs data on-chain they are doing that in spades. XRPL was built for payments and liquidity, I foresee this being the place where people will convert anything to anything because of a future massively deep AMM liquidity (not siloed into various L2 smart contracts but liquidity built into the protocol level). Adding XRPL for PRE would just tap into that for buying and selling and sending and receiving. You don’t need smart contracts for that use case. But you would need smart contract capabilities for DAO and self-custody staking etc. The bottom line is don’t become a maxi of any chain instead look to the future and what the project will need then start cross chain implementation with those. Cosmos good, Base ok, XRPL, Flare network, XDC, XLM, ALGO, CSPR (upgradable smart contracts, planned to onboard intellectual property on chain), TLOS, AVAX, etc.
PRE withdrawal restrictions relaxed further – The restrictions should not be in place for node staking as there is no way to game this. You might implement a cool down component to give transparency as to how much PRE may be coming on the market and to ensure withdrawal wallets are prepped for the potential demand. However, Search staking as it can be gamed should still have restrictions in place not sure how you can eliminate the gaming and extra non-organic searches, so I think this will always be a constant thorn-in-the-side issue until changed or eventually removed. But until then still probably needs checks in place before gaining full access to PRE. – If anything I think you should allow use of usage rewards on platform for any consumptive utilities before reducing access restrictions here. This would allow users who have earned usage rewards to use them for AI or consumptive keyword staking (if you change the keyword model as I have suggested) or other future consumptive utilities. This is better than giving them full access to the PRE incase some uses were later deemed fraudulent. In this way They don’t gain voting power or full owner of the PRE earned until proven legit. But like store credit if any restrictions are loosened it should be allowing those usage rewards to be used consumptively on platform.
PRE Base Layer 2 Bridge – same comment as Cosmos phase 1.